diff --git "a/data/train-00001.jsonl" "b/data/train-00001.jsonl" new file mode 100644--- /dev/null +++ "b/data/train-00001.jsonl" @@ -0,0 +1,5000 @@ +{"id": "train_05000", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9650cc91be34a06fab046edca6fa5fa339041471"} +{"id": "train_05001", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "documentation", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03545badabc72deddaaa02eca279212c8cb197bf"} +{"id": "train_05002", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b64e2d50c154deb66b16207a32cecfcdd7252bf3"} +{"id": "train_05003", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af102255028f89d052f4d8429e9784e3aa6eb932"} +{"id": "train_05004", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "documentation", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d3e7461e35d01a17637fb6e8257748c9ca61688"} +{"id": "train_05005", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tests_are_truth", "tooling", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "691e3f51693564664f1dce49db364981383e48bf"} +{"id": "train_05006", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6221f43a55a77ae8f9a9554ee8801e98d21a9a3c"} +{"id": "train_05007", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91bcd00b364ed312d5c82eea21ac8199838b08a3"} +{"id": "train_05008", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "governance", "reproducibility", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5d04c85560a48c645ef4e4fdc3681a90401c3ed"} +{"id": "train_05009", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "repo_scale_reasoning", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "896779aa8bc0725402b1ef746f8c4078313dce5f"} +{"id": "train_05010", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "reproducibility", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d995c71785a68b1fa204b0e2c74e73f8bfbf9695"} +{"id": "train_05011", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tooling", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be447c9a68b463a74dc6b2ed85d970e2f469aaa9"} +{"id": "train_05012", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "086bb3fff93504c68f9cc140c7416adeb01562de"} +{"id": "train_05013", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2789583a0b0bd35b30afced641088fb2c70258b"} +{"id": "train_05014", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "tooling", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "761c3821d4741e62b524a044821973fbb131bef6"} +{"id": "train_05015", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "auditability", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf84a303061d83155e3f75f9a28bf32d30f38e16"} +{"id": "train_05016", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6a2c58d6a090ee4abfb26b853d6781fe7704199"} +{"id": "train_05017", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe4199090669f8a1814263aa5152a8ae88a35c69"} +{"id": "train_05018", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d86a5caed88d40ed5bd4ae34c4b916198ec4ec2"} +{"id": "train_05019", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d8def0861a9907f767265c1076085995b79aeeb"} +{"id": "train_05020", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96d80731f3f93e943fb572501ad7284b618c73aa"} +{"id": "train_05021", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tooling", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "876edf2d507b385d12fada8857a46b58b888491d"} +{"id": "train_05022", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17629894c60819948693a10fc1f3c8f14f99ac9a"} +{"id": "train_05023", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b74eeafe7b6dfe0526f071fc04ea1e1e82be15e"} +{"id": "train_05024", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "ci_integration", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b0d9b8d4eed0e84de71cd0905dd04ae76de1b72"} +{"id": "train_05025", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "tooling", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f72bffdf6d43089278ebd63d7e6f8088002953ae"} +{"id": "train_05026", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "governance", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb622528085ff0131c5a9a87d726b98a3c9ae79a"} +{"id": "train_05027", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "ci_integration", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2fd1495f515b54ba3598099231f93f697bfa340"} +{"id": "train_05028", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "evaluation_metrics", "security_gates", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42b43ebfb3506368667e1f0abe23c7551a258817"} +{"id": "train_05029", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70a7cd6b8a42353aadf9d89ca753c88aa61e5e51"} +{"id": "train_05030", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afa49205c1a476389a973f90bc56e2875d6522fe"} +{"id": "train_05031", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba0447e2dc9dd3b56e4e1c3fb8a5ce499b1ebc36"} +{"id": "train_05032", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "reproducibility", "auditability", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f690eb62963fa5cd508bffe325d7d5392bbcf7f"} +{"id": "train_05033", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d3587d3deed4a1eea26a8b68521be2df9554419"} +{"id": "train_05034", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "auditability", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c91d6172e1d30d408e3e814dbef7e1634f2bf7e3"} +{"id": "train_05035", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "repo_scale_reasoning", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d9624ceaac84d4a9a40c2b791d4d7589f2a6b94"} +{"id": "train_05036", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "security_gates", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e8340b7c056f9fb55a075b9ed42859c5e966591"} +{"id": "train_05037", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed17bfcd85c57d11a8efb72967e24daa976a0886"} +{"id": "train_05038", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ec6bda08136c38d9e74e225ede5112ce7dd556e"} +{"id": "train_05039", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f22269b0f4c1c98cf0df1e137656a75be7693b5"} +{"id": "train_05040", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9a1d71ea8f44dbf8b6b81a611809fcc5243f2ca"} +{"id": "train_05041", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "documentation", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54f90d9ff562a8071946763fde64d042fc0fac87"} +{"id": "train_05042", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "824b5ba2fee0ce88852eb2999f227db685482b46"} +{"id": "train_05043", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "ci_integration", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3032205c6def0d1b97243fda49547ac462997167"} +{"id": "train_05044", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c97ed724e93ef6115d5b1d88c26cab73e5b1c09"} +{"id": "train_05045", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "ci_integration", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9cc98d4ba9eb4a6ceff99e96b76ef0be6a0e64d"} +{"id": "train_05046", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "461ebce493bf5c7ebd94dd5f0a3a90ed0b4801d9"} +{"id": "train_05047", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b6f6a435eae6f316089115becfdb04618cc4dfb"} +{"id": "train_05048", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "auditability", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b378c3960a8c7909d44ed26061d6952997f5125b"} +{"id": "train_05049", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fda98e779778c58106195de5f80b0db7358ac165"} +{"id": "train_05050", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "tooling", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0280a259946f27352a901bba2910ee435ae4e8bf"} +{"id": "train_05051", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "468755ddf7ace12722fadbd528ba44d81017cd6b"} +{"id": "train_05052", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "641df3573b0ccddfdb8f29c2285e28af48dfba5a"} +{"id": "train_05053", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9dfbe6aa8ae7478c185dd00357ad6282561c0552"} +{"id": "train_05054", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39aa1bea7a45d8d150801e87b6b38426c4e2e0e5"} +{"id": "train_05055", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a4290c70fd8b825cbd390870e0dec1095c4eb99"} +{"id": "train_05056", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "893a5fc22bf0fbd67e23bd01cc32733005cbb8d6"} +{"id": "train_05057", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "documentation", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2abb516b785d0a08d483043a9606fe7983362a4"} +{"id": "train_05058", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42959c550d9f036ecb960697940664130e7ad9f9"} +{"id": "train_05059", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a205d20ec4e2949a55417dadb32e05e6dfe14017"} +{"id": "train_05060", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e8c485f1168a3fb24b0696c7087d6ec75e9daaa"} +{"id": "train_05061", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "auditability", "tooling", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "808ff9396fb9f5d293e898bfc40aca343107b3dd"} +{"id": "train_05062", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6dcc297af131f6a8a47c3c6105eb150484e8ff96"} +{"id": "train_05063", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87e3c823f8b10d5d6c91bfb5007355f2c25545ac"} +{"id": "train_05064", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fc480daed913ddbe584c446fb9c6c6717420c37"} +{"id": "train_05065", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "tooling", "documentation", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f346aaf08f021fb592e01985fae800c81cf4fbf9"} +{"id": "train_05066", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "reproducibility", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71aa36d01fd476d061e65a302346673a07f3495a"} +{"id": "train_05067", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f80196915192b409205d8dd63f83d40a641e9b2"} +{"id": "train_05068", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd2c361c94acc4ae5eecb2464b855231b56da070"} +{"id": "train_05069", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64697bf40bd97dab151d6ece41c16e1932128c8d"} +{"id": "train_05070", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eed67ba6cb3af11e05409f18b290cc43f95c74be"} +{"id": "train_05071", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a56c8f91ca5dc2a8cf28cce3e951ba1e7c8b92b"} +{"id": "train_05072", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13c654f63a68559362d7f3d4fd8bc1c429ecf9a6"} +{"id": "train_05073", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "auditability", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f2bc9308c1a03ad72740b4ad17db017be9a80f9"} +{"id": "train_05074", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3eb24916c676182d390dfbc9c3d30e7d85a87c7"} +{"id": "train_05075", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "auditability", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4538c35740dda6c0e9116acc48767280d21b489e"} +{"id": "train_05076", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc2d5a3c63c6f5ad0af2edf3269a5282ff6a5800"} +{"id": "train_05077", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "187b90fe7856efa2e5757a45818c17b674b52a1f"} +{"id": "train_05078", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03faa0bc234fd8de50fcdde9518c4538826b5d72"} +{"id": "train_05079", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tests_are_truth", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6268cc4f1d8cae312b2534aeb426b5dd24bc434e"} +{"id": "train_05080", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13159a0bfdac9b0e0dd89b1d8f16bbc26649c7d1"} +{"id": "train_05081", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d564759775a0479a8ce9c45234b14f8e627b27dd"} +{"id": "train_05082", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1546dc3dddf616c996131e5d44e206c9bfaf6c5"} +{"id": "train_05083", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "401534976bf96bd7f26be813b71e71e976bab789"} +{"id": "train_05084", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f597371d99a821b0eed5f3b5a2035ee5da5b5b09"} +{"id": "train_05085", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7632bca8831d1b73b293b27090ab30705ff75683"} +{"id": "train_05086", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53e1d01d7d9396b05698f26035c6ab3c9584772d"} +{"id": "train_05087", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a25229f17f4afb874eb97fc3940c2a9a06b86683"} +{"id": "train_05088", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5fdbed797bc051134daaf3254bcef9e137a2731"} +{"id": "train_05089", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47d96750360083eb693a6908569ce2c4206bc73b"} +{"id": "train_05090", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51e0d8e56bce643faeaa51670ad6377293168503"} +{"id": "train_05091", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a697976e2f02f6100bec23286d2bfb601efebe65"} +{"id": "train_05092", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef743eb10e5038d33f26555e2ee1aa0cb249b2b1"} +{"id": "train_05093", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b83f8401a8a702ec016bc4c5eea2d2b852959ff"} +{"id": "train_05094", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "reproducibility", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e4adde41f0f0b0bf139345dd64f0f0d1db956c1"} +{"id": "train_05095", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "repo_scale_reasoning", "governance", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46acb660a144f8092a17ea0aa0f8d63bc75e40d8"} +{"id": "train_05096", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d39b4bd88d83cbc9e80f64748e3798d81e17ea21"} +{"id": "train_05097", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19dc817ce69bb058b5d473a7e162d04d1eba1c0c"} +{"id": "train_05098", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61a7ead3504ffe27225c374f9246ec4050cb6826"} +{"id": "train_05099", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d846669d081406c518a27b9904d0cfcd6dcfbdc"} +{"id": "train_05100", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "evaluation_metrics", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76b9174e991395f2be3ecee62a6b0944848bfce9"} +{"id": "train_05101", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9973b1a1002ffbc31af60fe0e8aa0acc9a66585b"} +{"id": "train_05102", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4032351d0072bf16bf14b22270144b522d32b0e5"} +{"id": "train_05103", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "documentation", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f22fdb91d9aca449debd540770e3d69ad91332e4"} +{"id": "train_05104", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d63653196c1c75ad34215f1ca7796b9a86fe219e"} +{"id": "train_05105", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "702dbf27f2f2325b62e727804893f526222944ee"} +{"id": "train_05106", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "123a195a40e7c7159df7daba3511ba04251982e0"} +{"id": "train_05107", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "security_gates", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cbdbb46984c0489627834d9b17d0adfefb30900"} +{"id": "train_05108", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "154417faebc9fe7206ecc1548178c76735ec0dc0"} +{"id": "train_05109", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63b42f02d5f739ea7b2138b6ae154ce78b6b8a7c"} +{"id": "train_05110", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tooling", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd2ca39e588e9a7aa27fc2268acf6692963de60c"} +{"id": "train_05111", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f73b4c022a3b1b99def755fd74a2562a466e7f9"} +{"id": "train_05112", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ded5c643a9d74d88f7eac911892c28d7a41fae50"} +{"id": "train_05113", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87c6a2b977b8c8173193b735032dd35e3f240c7b"} +{"id": "train_05114", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "343f2b97f8d5d6007e715c256347c0e15120fc07"} +{"id": "train_05115", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "governance", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc118fcf2bc4ce0522c174b4aa0060e9ff95331d"} +{"id": "train_05116", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f31a77224ddde94109b913933f07a53eb67e0ee2"} +{"id": "train_05117", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "ci_integration", "governance", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e9f3dbead09f276daa3c5a9bdfd1e3964f21d08"} +{"id": "train_05118", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c1451c9d1e91b34cec658e06d3c383046f91736"} +{"id": "train_05119", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c4f5469c81337a271999f5c56c0d3e4bfc5439c"} +{"id": "train_05120", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "reproducibility", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e228b43f1e7d6478a6ed99df7ed8941b2fa0071e"} +{"id": "train_05121", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8b2a1fd9eca68baec40ec31ee4b176573e66c74"} +{"id": "train_05122", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc23fda4c2757e055b488fa7010c63fd2b72189c"} +{"id": "train_05123", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc768fbfa0ec2878e2c7e1d30bc74a22312fd6ee"} +{"id": "train_05124", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "security_gates", "reproducibility", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a459e0775993575a9cc09cd42ec001cf831b2a23"} +{"id": "train_05125", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3a46cc7ac498e2f1792f0e3a30fce270e5bb6cb"} +{"id": "train_05126", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68ee3447de9bd7b5cd6597c77c05ee1a70c50e1a"} +{"id": "train_05127", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0eef96e56c254971a405401645dfa6abd1e132de"} +{"id": "train_05128", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "documentation", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f496a7a88e6055717f6aed9b2cbf4b864a3d9e6"} +{"id": "train_05129", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "ci_integration", "documentation", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ae0b7025856f69407a1c6869e65eb44485d8933"} +{"id": "train_05130", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7deeafb6097c3bb3532dafaa9336412e23e1705"} +{"id": "train_05131", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ba069f0a55e73fe462b825ba88bc6094404266e"} +{"id": "train_05132", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b001c19694757e4b443040f1231fb7b7db912e99"} +{"id": "train_05133", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "tooling", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd933c65ebd32fe3e05f660c4a8e8bf87ca3ae73"} +{"id": "train_05134", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "ci_integration", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9aee70db563c17a8af40a1c4bdb39a0dbd89023"} +{"id": "train_05135", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2016148a7bb1a6574eddf1f0b1cfcc26bb3e3442"} +{"id": "train_05136", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "documentation", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c7291cffebd3c9cfd437f3b1dd88835a33f91ea"} +{"id": "train_05137", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a156d4bb5ebd41fc1c8627fef39a764adb866a8"} +{"id": "train_05138", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "evaluation_metrics", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "787129152f7be2c0beb93db10e5b30a32295163c"} +{"id": "train_05139", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fbe83735bac342372d7aa79eee94357f38790f9f"} +{"id": "train_05140", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31f172c546094c38f332fc1308c5f0ab78e9c7e5"} +{"id": "train_05141", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "ci_integration", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08b58c5351a97153d6e77d156a157dca59c94dbd"} +{"id": "train_05142", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f07fde52bd698111dd6d601868c19302d7d18a72"} +{"id": "train_05143", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c29a0db196dbc38c3e4ef0f4f2b914f9c066865"} +{"id": "train_05144", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd1a692dfe8610ffd9885e5a616426249f4c33fb"} +{"id": "train_05145", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "reproducibility", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ca76874698a696329d9fa1ecbeb1273d417e9e5"} +{"id": "train_05146", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90425b9bb2887d3df59e98d345d3fa06bd3599ac"} +{"id": "train_05147", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "documentation", "auditability", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b96a8f42700318886c27ef50bd97877328a11002"} +{"id": "train_05148", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22184ebe96d6a0b262273756135b9cfcb3e401e0"} +{"id": "train_05149", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b95b94290afaa00f27e7313edf4e481fba9d46e8"} +{"id": "train_05150", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54b10f94ca99902ca1e3a4e3578090d97dd080e6"} +{"id": "train_05151", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7cfef243890c2de26e3cf07b446ef32ebe8a0c1"} +{"id": "train_05152", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5671f80633f13db5401223bd74652b73bbedcf61"} +{"id": "train_05153", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "security_gates", "tests_are_truth", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c55635a4b360f2881d3b33f812782048c239f568"} +{"id": "train_05154", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7b902b0843079348a76cf5d1d0e325bbd71274e"} +{"id": "train_05155", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a21e406d6e5741ccf08b709070207f2108d66bdd"} +{"id": "train_05156", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7028cd1ac431280aaa605e0cbd6d42a78a014dac"} +{"id": "train_05157", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "reproducibility", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69ec45c1bfc605eb3814afc4959c8a36121b708e"} +{"id": "train_05158", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04f7b268cca04654060580bbc1ed6dd1416ec10b"} +{"id": "train_05159", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3511b4cfabec8ca62ffe274c2263ed6788512c99"} +{"id": "train_05160", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6abb2a0a3be4e5eaa1a0bae1cdee4f91798dc6f4"} +{"id": "train_05161", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tooling", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e05aab42546efde141269e154397ebe7f90ab9e"} +{"id": "train_05162", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "984f53c22d4ef8e5c04fc286f29ad20324584a62"} +{"id": "train_05163", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09ad791e9bc113fe7c16e91540b88421fd38c8a2"} +{"id": "train_05164", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3768d71cd78f2e0cd11b4c8f85d513cf75b45017"} +{"id": "train_05165", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "security_gates", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd5e8cbc7146e84fb19abcaf621371b94575f0de"} +{"id": "train_05166", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "334d1fb166fda9b33e830fba9913ab899c217213"} +{"id": "train_05167", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tests_are_truth", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "889b8de681d5bd7ac8a66830f09e3a3bf6acbdb8"} +{"id": "train_05168", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b90a38cd5aa10fa9d4af00339667f9d2c2965bae"} +{"id": "train_05169", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a4ebd2f2b2cf3c09c61e1cee6d53a50545be76b"} +{"id": "train_05170", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "ci_integration", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ddf2f41c5b1bef16a6a862b8f8452edd611f618"} +{"id": "train_05171", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3554b25225410faceffe0a9cd53db5cbd931fa8"} +{"id": "train_05172", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "102737d0d26a3a6ff789059a31299546a73f4cf0"} +{"id": "train_05173", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "documentation", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d184e4635c1c650cb3f959431bc1f87dcfd20f4b"} +{"id": "train_05174", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tooling", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30b53100ce0eb85c6afc1ee9f1faff1ad4134cbf"} +{"id": "train_05175", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2eb2aed57df9c6c5cb588142bca60ac1aee2f53"} +{"id": "train_05176", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2d66175c49d98f2491e806b365fd77ff0662b24"} +{"id": "train_05177", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "auditability", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "292918419c58b1b7d184f0a3d3e9475547387471"} +{"id": "train_05178", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "documentation", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f50791261ec1039acb37781abf6749a13d1dbfcb"} +{"id": "train_05179", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a62f5f7959731cdbc19f1d23bb1de06a47fd8be3"} +{"id": "train_05180", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dde7913c23b14596decebb862ce8e3fc0e4c5eb"} +{"id": "train_05181", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "reproducibility", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cca13bcf6835e2cbf3d86a32388453c454f7dd16"} +{"id": "train_05182", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2464bab7457284c77da8b573f4f25d120e9fd1d0"} +{"id": "train_05183", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9e4e0faf7d4ec8c821f2febfe3d100e16af3b3b"} +{"id": "train_05184", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "ci_integration", "tooling", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65c5c4bd22654a2923f38a09babe0bd0b36e7a22"} +{"id": "train_05185", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7259a77f24917652d42abf5f061df942d08a6d3"} +{"id": "train_05186", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1c2040af1f2e8b03d313ec64c4be68d60bc2056"} +{"id": "train_05187", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "security_gates", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e1325da28c8387c75123d690fac7c1f54601f6c"} +{"id": "train_05188", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34ff7256f33e8c6f82f43461efc3a03698383dfb"} +{"id": "train_05189", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "evaluation_metrics", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87640cb25da76df6692db20c671133b5ce60af1a"} +{"id": "train_05190", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "repo_scale_reasoning", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5013e8b89212d73c9462da9d2f93299d317bbd0a"} +{"id": "train_05191", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "governance", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49c7b2b73ab0073fe8c9e6d046a166a2bdf00d60"} +{"id": "train_05192", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76af7c5f84dc9244bbb9c45acbeb8a893b79067b"} +{"id": "train_05193", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "851555e0ff602a2d47a88fd61447844c33ffd3dd"} +{"id": "train_05194", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "reproducibility", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "170306f94467c6f1223272f7b35942e7d08fe3c0"} +{"id": "train_05195", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0b73e1c4722443f07da0f5ecbd5d1dd9371ca7e"} +{"id": "train_05196", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c86c24b6ba5249921a1ccef71ba7703a74ab8f44"} +{"id": "train_05197", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf0499219bdc82449d26c3affb0e9d837fa9bbdc"} +{"id": "train_05198", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "security_gates", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0a7cd9730e8fc0039ea8f495799dbf3cfeb519a"} +{"id": "train_05199", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8995dcc1ee03419e336fde101587245c7ee402ab"} +{"id": "train_05200", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f417b35c25328b9f4857d2e5cae0f48498903570"} +{"id": "train_05201", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a41236f227c62de0f2755257f704a22469a32790"} +{"id": "train_05202", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b214b97fca46a3628cf3847e3d2f1ffbd147f5c"} +{"id": "train_05203", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07154649c872c9b6ad13a62b9107e23cd06cc955"} +{"id": "train_05204", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89580e238df2034ab23fb67305313727dcdcf58c"} +{"id": "train_05205", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3371a530bdecf067d5c38f542b83257f04a0d339"} +{"id": "train_05206", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tooling", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c0b05adc88f8ad218c6eb0b4c31177f5bf2b056"} +{"id": "train_05207", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "tests_are_truth", "ci_integration", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecafa47f8cef0ff589d6019130d68102146dcb59"} +{"id": "train_05208", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6945b99adaeefaff7a0ee96224c7f38853bae2ba"} +{"id": "train_05209", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "evaluation_metrics", "documentation", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9a5fa06f23e65e106deeace7f7937750324d555"} +{"id": "train_05210", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "tooling", "security_gates", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c476d147ead669d64679c8cd2aa66f554291abb6"} +{"id": "train_05211", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36748c22f2c8c34fe1affd7d9b5a7e23d0982dac"} +{"id": "train_05212", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "auditability", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51fe12ec33ba690b357503d4064392ce97e11145"} +{"id": "train_05213", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c8c9e9820c3bb745305955d13ae9b9c92a50136"} +{"id": "train_05214", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca11ce9e437a7eec1d7e71aff6bf30cbff9ac436"} +{"id": "train_05215", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tooling", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5359c5a12c83dcc3773a74e6c194b3adac79f886"} +{"id": "train_05216", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "tooling", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b75788aba9ea859cf9df25bca91594fa8bcb21e7"} +{"id": "train_05217", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1454f199bf24602ef7f301df360423c031b7ce0"} +{"id": "train_05218", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "ci_integration", "documentation", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cde3c74f962e58cf7baaa8d4539ce2c1f38a5c47"} +{"id": "train_05219", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75f4cbf0493487d28fb09bcec576ab559ae915e7"} +{"id": "train_05220", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad3dc7aa518a48eb9130b35f77f3440b739a237d"} +{"id": "train_05221", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ea4e1b71082ea0b552971398c08ac6cb4075d8f"} +{"id": "train_05222", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fb9a456c606dfd47c017c012e15a1e4d628123b"} +{"id": "train_05223", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e21da65bf175ab076a3cf2e57bd39cda945b5b6"} +{"id": "train_05224", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "tests_are_truth", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd96dbc8794301311c55f0f7a3260b2861cd4065"} +{"id": "train_05225", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "reproducibility", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ce0c15f3b3fae3139251a3c20e6de4cfb7fdbaf"} +{"id": "train_05226", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "auditability", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a389d99b18d941231b683d198dcfce8e9d3274a"} +{"id": "train_05227", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b5b85f5054deaee8de11d7dd098cb0069066d88"} +{"id": "train_05228", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "auditability", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da670112a6b6774c45dfe1ed054d31499237affc"} +{"id": "train_05229", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "444820161284805e1b710cc12057aa29537834cc"} +{"id": "train_05230", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "531362f94f561efbeb610c4986aac5bc28a61d41"} +{"id": "train_05231", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2f67df1e60b9747a1fb43be9321e6abc5bebe0d"} +{"id": "train_05232", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daa3a123d0c47d278c529517f3f5763a156a0c1b"} +{"id": "train_05233", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tooling", "governance", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb5bf7f8867703de9e14bec9692227ac3bb6d7ae"} +{"id": "train_05234", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0bc40701a8450f1b0d8eee458b22a930aa4f112"} +{"id": "train_05235", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64514e82b2225fbc8ddebc8a738a55e84a0b2664"} +{"id": "train_05236", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b09cb3531598bfe97b45c0c43d4b4f9e1be8798d"} +{"id": "train_05237", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a9f4ded54b8fa7c1ca1a13da071b7cbf54e313a"} +{"id": "train_05238", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7c799196d99d8e2d50262adb0993ed203c15623"} +{"id": "train_05239", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "governance", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1deaf8938c0e1ab3455f84ea54367ffe232bb5b0"} +{"id": "train_05240", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "evaluation_metrics", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4876660cf5bb16414b3d9d6c62032aae99de79c7"} +{"id": "train_05241", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58db8de21105f42b6f7a69ba7e2253d275683f38"} +{"id": "train_05242", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "documentation", "governance", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68df4ff8a09d9b3d2866aec98c58b1a51b9b81e3"} +{"id": "train_05243", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "governance", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e80ccf7f3c394874e94a2dc7f921ea43717dfb39"} +{"id": "train_05244", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "459fa43b6486e1e149c796025265d889b8945fbf"} +{"id": "train_05245", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cb5765a64c33d3a00a90f56bc5a23049b946854"} +{"id": "train_05246", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b3d56992a2df4e7d9100fc037f8e471786bdbca"} +{"id": "train_05247", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca1d850773358e9289832df11f5a0aed3fc1d026"} +{"id": "train_05248", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64d47272dec5dbadd197ad3df2ea7084f50f1193"} +{"id": "train_05249", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "auditability", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbce10eaf65628d0a0a6c6c9317611a80f3b5e45"} +{"id": "train_05250", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bbe159e5d01f65953eae3e6146f6874297b3f2b"} +{"id": "train_05251", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "288371d5c3d08b43ed100d40932501bf4ed194ff"} +{"id": "train_05252", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e257b5d4a83d9a61a30a5b00042ab3ea79cd0c3e"} +{"id": "train_05253", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f59a78cdda57663dd94d08be69bf0907ab7b4115"} +{"id": "train_05254", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e93fc0e9b0b17daddcd12a63e4429db5bf7043e"} +{"id": "train_05255", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0552cf7459cc1f45180a21d2c90ec50874e0080"} +{"id": "train_05256", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tests_are_truth", "security_gates", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "011ca5431e7a6c5f8b7801361640b6000db0a035"} +{"id": "train_05257", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bff0b824301a5c0b4089678ea854f10cbe85b5f7"} +{"id": "train_05258", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9edc197a5c484c10e75cc190f95afee2868d12bd"} +{"id": "train_05259", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f32954c129123566aa8f81a94dd4d005eb61f67f"} +{"id": "train_05260", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77dd33830800edf6138b0fee69b5592c3d03701a"} +{"id": "train_05261", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efb37294514038a1167ad04ddfcb5737009f29ee"} +{"id": "train_05262", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ded5c643a9d74d88f7eac911892c28d7a41fae50"} +{"id": "train_05263", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17893aa97dcb6f570b08b0d6cbb2cb451b0dfc4c"} +{"id": "train_05264", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7531a5a7be2bd2420e163d2853eb77df1c1cfa28"} +{"id": "train_05265", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "governance", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd1817a67c588ac4c8a16b78b708b3fce83a27c3"} +{"id": "train_05266", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09b57087d9277013cdbc595bdd9b089a1cb1fce8"} +{"id": "train_05267", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38dd9970493480e9438f95a0c842a2ec48ee6678"} +{"id": "train_05268", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd173b0ca2d9488fae2105fbf12c400308d28312"} +{"id": "train_05269", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "documentation", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b38df81d6e323c7b852da2a969ea859e5cdfb8c"} +{"id": "train_05270", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "238640525c7ffa35379d102c2e177225b5809756"} +{"id": "train_05271", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30a471505b238c15df03d5673d3cbec08fac2824"} +{"id": "train_05272", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "reproducibility", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "296427a06c2bb7a98432afc9bfb73bc75ea083f3"} +{"id": "train_05273", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54e57c9a53644ba59b9e4fd596ba26b62dc955c0"} +{"id": "train_05274", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f88e362f58b587178fafb19813368a049eeb63e0"} +{"id": "train_05275", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "auditability", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42b4a1e154c78beb81db1fe6e2c3591759cf196b"} +{"id": "train_05276", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "reproducibility", "governance", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab75663ac324d21f43c5a5e885fb9d8ae99375ba"} +{"id": "train_05277", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecd8823a69361f27b2ff4bb6eaa6d45a94e88354"} +{"id": "train_05278", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb32dc2924139868ae3b62b26b5144b7a11c56f6"} +{"id": "train_05279", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd1fc3b2405e8204299b1a368af34332d39fa716"} +{"id": "train_05280", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "585670101266adb4dfa3360a0f001299935fc190"} +{"id": "train_05281", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b24f9965bd65db1bf49f10924311e3cf8419fbe"} +{"id": "train_05282", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fec7354cca3cb3b54b0a5bc0c8943ebaf42b9ba"} +{"id": "train_05283", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee6ebe35d000574c895e6e02e0031f53a10f615a"} +{"id": "train_05284", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed9e5c4a332683b398a449b6ae543989dac54635"} +{"id": "train_05285", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0466e9757cb7a0b90695e120bfc8e89ca6644d3f"} +{"id": "train_05286", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d1b79ce69f0976c0902477a48137372b026cc94"} +{"id": "train_05287", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef16b6adefc251cfb38b40e8adca697440988243"} +{"id": "train_05288", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc992f4ee6e283e3a7f2330341b1a304ca9fe148"} +{"id": "train_05289", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6793f5bf0c084301ddac384f1d3bc23215cd6b2e"} +{"id": "train_05290", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "documentation", "governance", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1626ca14dc64c6654bbc3cb6db34130d0512417c"} +{"id": "train_05291", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5304a7eeefc19c1f379a537780fb749de929bc91"} +{"id": "train_05292", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "auditability", "governance", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81080fa338068f92d11f84e85fb37de4899cd970"} +{"id": "train_05293", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a13d398f4c4bda3e5bcedf763b840212ccda023"} +{"id": "train_05294", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "ci_integration", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5861c44ae0af941b68f3e713b5ea7576a7a60210"} +{"id": "train_05295", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6438cbfff1ae17638b7e3ceb7c244d7ef318f3f5"} +{"id": "train_05296", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "reproducibility", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a40cb459224ae5640510babf1701ab8fda9c7841"} +{"id": "train_05297", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d2b3e4b4ffa9c07f7af81ee9bd82730b08cd978"} +{"id": "train_05298", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "documentation", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bced7b13748bee5ef8f83d47609f6153a10ae6bf"} +{"id": "train_05299", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ac3dc5c930cba55349bd68e8859e9d5f498b17e"} +{"id": "train_05300", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59fefff0062fec7532a5f191e8051c57e5cf900f"} +{"id": "train_05301", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b99750905cdb0218a56f06c6905c65ce229e473"} +{"id": "train_05302", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tests_are_truth", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83ec8782a39621696f53fce70587aef5834e90f5"} +{"id": "train_05303", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tests_are_truth", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49227bfe001125da79462a07f085ef0014fbc023"} +{"id": "train_05304", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe713ab07e7387753db0ff0948c87090e58e154a"} +{"id": "train_05305", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "106468964a9a96fe9b1b450fa03ace535c089d5e"} +{"id": "train_05306", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b75bb1b46198e027f76b0051643412fe10e66704"} +{"id": "train_05307", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73ba525496d7f5e7d21da1092dc4e0c5f33746e1"} +{"id": "train_05308", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "ci_integration", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87a2584c291312b62ba35ff1a91371ea72b4d1a5"} +{"id": "train_05309", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "929446e2c15f1ad2acc65c858e7bde748b4c911e"} +{"id": "train_05310", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "254b093138b9a947809e21767e40165814e8a13c"} +{"id": "train_05311", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "documentation", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e312535c26f2780917f94d6420b435e2e06eba7"} +{"id": "train_05312", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "repo_scale_reasoning", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dddf6f8cb68e0bb837a910298007933221d9fce0"} +{"id": "train_05313", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b44c09964d2e7c43c067f4754170e145f459dd5"} +{"id": "train_05314", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee3338ebf48e4cefc9733bd16564de2b002c701c"} +{"id": "train_05315", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "auditability", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef9db386931de591738c1d6959bdbd6f01502ad3"} +{"id": "train_05316", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fa02caebb88c87865a440c622a68fd3829e9368"} +{"id": "train_05317", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f4319e8979d51d95793941c280f0eb81b267e0a"} +{"id": "train_05318", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "governance", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e44e8e5280e067d801d7126035271cad15d0afbf"} +{"id": "train_05319", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93648707421981306dcc718720c0818cc51ff154"} +{"id": "train_05320", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "reproducibility", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9313a09193b44ad34f4d9b7fd0a9999b6960c147"} +{"id": "train_05321", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da113b3e85968af6784fa0bda6169616e06b3dbd"} +{"id": "train_05322", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "reproducibility", "tooling", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04cffffa32f25b726a1cd0d80785899dbb9a87af"} +{"id": "train_05323", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "documentation", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3980d445378945e0e5c99164d2f940c5d854562"} +{"id": "train_05324", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "tests_are_truth", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "956e8b4b27e4b2f148ff1aaba14d1ce204b05865"} +{"id": "train_05325", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tooling", "security_gates", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68ee3447de9bd7b5cd6597c77c05ee1a70c50e1a"} +{"id": "train_05326", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tooling", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2108a323cc3375301b159e2d0b936fea98302fc5"} +{"id": "train_05327", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89b8b3ef719f2c33f645e3aeede2ef4d082abf85"} +{"id": "train_05328", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10a2264d0cdc22ec02519dffce335494b309dc70"} +{"id": "train_05329", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "007042f515a3ddf82139884b1a32df200c0f0ae1"} +{"id": "train_05330", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "335384b09d3920f56cd53884edd5ac4a5ccc8af0"} +{"id": "train_05331", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "documentation", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aba91ff88e5904f4b9e45c44531d0a57f45ef268"} +{"id": "train_05332", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4970e35b3af102327ecbcf4f6fe4323cfd87405"} +{"id": "train_05333", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "reproducibility", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eaadc7013d6bace09e99e2af274da91ebcb48405"} +{"id": "train_05334", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0112217826ed4da398de441efdd7660c868149da"} +{"id": "train_05335", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95e277a13790952585f626297d6e2d054a131d2a"} +{"id": "train_05336", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "tooling", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "283b125fe08e03d282efe1e2f83a9b6dcd4e40f7"} +{"id": "train_05337", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da9c6033cc5725cb5c7934c92aadc2091221ca3c"} +{"id": "train_05338", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "reproducibility", "security_gates", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26d3be2a96b96c7cd326c1c114f35849b111c808"} +{"id": "train_05339", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "591fb956552f22f601423eadf4872803cd4299f5"} +{"id": "train_05340", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7c547f6608ecb3663afaa2fdb5ef1fc59fbff21"} +{"id": "train_05341", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c071297246d610823e91dc2a5ff15bb6ab99832"} +{"id": "train_05342", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f4d118f4c360e1f909612afd21fb1088e674bb8"} +{"id": "train_05343", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c5f55a57f6a3b06ae82850bfadcbe3e69fda95a"} +{"id": "train_05344", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a6fc290633ee6e79c00f06e43776c2c4ad5b936"} +{"id": "train_05345", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e34b639bcb13001c0cc23ed4807046d0bec98433"} +{"id": "train_05346", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69982933dc221e5574533219d415f8bbbed458d0"} +{"id": "train_05347", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "tooling", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9fc24e59f14b9f209c8adc77933fb8a945e2773"} +{"id": "train_05348", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "160dfd1191f2ebc194bc8d220c01d94ffcd4ede3"} +{"id": "train_05349", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b61078a9336a81f3caac3e6479f873e2b8a79547"} +{"id": "train_05350", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ced1e24607896ff38ee40cb01f3249692569faf8"} +{"id": "train_05351", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b127fafbc9312266f3aafa5c15b1466f25a8fed"} +{"id": "train_05352", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "evaluation_metrics", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b967e40f25e2d6b861ee0516941d8413d878548a"} +{"id": "train_05353", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tooling", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b65097b6beaeca96551ecbd747fc5b8af0a59e2"} +{"id": "train_05354", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13e24033c7358a61fb54ccff8340dde8cc07871a"} +{"id": "train_05355", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "ci_integration", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "270bb1d3976f9e31589497e5dce1dd8751a82818"} +{"id": "train_05356", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56428b3d45c85227a1554df63d90c0e4b4712829"} +{"id": "train_05357", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "679a23971f4c5aed20525577296965349bd1c4f5"} +{"id": "train_05358", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20a05f08aaa8fc44d65d041918d65c2063ed57ca"} +{"id": "train_05359", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "governance", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef83aae2009311eaadf0e4e5342ccdf20c30c6ff"} +{"id": "train_05360", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8acf60ef7079e29fc7b45060db56bccc6de2304"} +{"id": "train_05361", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "ci_integration", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a07183ef82d20c549ccbfbf3b67a35bddf85282"} +{"id": "train_05362", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "ci_integration", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19fe0f03f14f31aa0c8e6e3b04d6d3d94bf23535"} +{"id": "train_05363", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tests_are_truth", "tooling", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c68cf98876266a70de2a6c89939280aac4c9737d"} +{"id": "train_05364", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8e9ff942b26b9dde770c8d0f8f298e76468183f"} +{"id": "train_05365", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "security_gates", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea481c98dd4f98a52cfcf20886cf4c64d8842bcb"} +{"id": "train_05366", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f1a4dfb89d6a4ce3eb5a26917baffb96c34d48f"} +{"id": "train_05367", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7138ee299dfce85d6d84345348235f73688d1bdc"} +{"id": "train_05368", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4afc1d045c3ea92ac9253c23de5ba37618007212"} +{"id": "train_05369", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "baca87c0da1b3e86015a915c9b23ce5aa518d45f"} +{"id": "train_05370", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62f0bcad36891d67012cea4a580bb2f157419ca0"} +{"id": "train_05371", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tests_are_truth", "governance", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41aeac108cf06fd99db80cad36e737824d26af98"} +{"id": "train_05372", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6d836d1170301c2274c31dbc37e8a4540c99697"} +{"id": "train_05373", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ee1e0b444da60df9208d10b361605b76cc573b9"} +{"id": "train_05374", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "documentation", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5ed8baa86bbf24cf575eb40a92f1075e94d4a19"} +{"id": "train_05375", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bea19cacd1e534211c724927369db9842bed2c22"} +{"id": "train_05376", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1992fd9829eb2d9fcf7c6e3ae9c35cd7499bf512"} +{"id": "train_05377", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a58e065cee175022e7bf9c4a9510235c517cd881"} +{"id": "train_05378", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "ci_integration", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd6843281ee952a156ffe34cb9de205d66a5ed76"} +{"id": "train_05379", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7935f4b064118e62a7e75d63ad5eb35d19b94c32"} +{"id": "train_05380", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2003591781cf3aa13cc8b7958ea7292b8c65ced"} +{"id": "train_05381", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51b9b2afb1794e548be54b3920df58bc4b9787d5"} +{"id": "train_05382", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5daff8397bbba8b1341919348eaa0446f801277"} +{"id": "train_05383", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fceb468c8cdbad5eff159f986944f72daf90715a"} +{"id": "train_05384", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "reproducibility", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd7ec16e575433047ebc4c5212f4f7590967c779"} +{"id": "train_05385", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0d54473b49d84c6147d7056073e6b2111933156"} +{"id": "train_05386", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4fd83a756d23be0ea909dd7efe40f3af4abaa8f"} +{"id": "train_05387", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc4ce30b838e090f53809a7350a0b1324b7ba1d5"} +{"id": "train_05388", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f04e0f6205d42a2dac1103baf7451939fbd45877"} +{"id": "train_05389", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07a35cb0a6d9b7882d37ed7a76be4bcdbc8f955c"} +{"id": "train_05390", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "963702baf8d86ab6fb3adea400b82ef0cc2204bf"} +{"id": "train_05391", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "823e35fa0431c38a2c8d4b867b528edca25490f7"} +{"id": "train_05392", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10285ebbee84b838594f961e2c6d02f4a55a7f4a"} +{"id": "train_05393", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2174ab0f2f80612f814a45b8575dd10d215cf34c"} +{"id": "train_05394", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c65495912986049712578eba468f0bafb29fc40"} +{"id": "train_05395", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f104700d81732069f945b8a9e01f460e8818b609"} +{"id": "train_05396", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a914947c923543b8099bc5a56769a5170ca7ee01"} +{"id": "train_05397", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ed42e96c8b5ff8be4373d13d5dbb4dc65a405b8"} +{"id": "train_05398", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0d00e7b1fb0a318f4208134ef1c70922df0b52a"} +{"id": "train_05399", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "auditability", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b013bbb18c29d13436738b7720fd53bdca30b2c0"} +{"id": "train_05400", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c469f5469aeb28a4c53e90b0e02a948649aa5b4"} +{"id": "train_05401", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "documentation", "governance", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68f43362a197570d34b61828b5aca64380d3c935"} +{"id": "train_05402", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "repo_scale_reasoning", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d58c0a4006eb35d911fee88957a700321e9b1944"} +{"id": "train_05403", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "546e1744ad7444806b8225d5d4ed4e925fb0f080"} +{"id": "train_05404", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "evaluation_metrics", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "509acf19c89aa304ca560ba8f44ff4663848223d"} +{"id": "train_05405", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53f12c8e561269b168b877e735a05266430aafb1"} +{"id": "train_05406", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e2f68f62f5779c0bb5397631585ec27637a80b4"} +{"id": "train_05407", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be71a7589ddae1839a5244c0e194654f68f9f8c2"} +{"id": "train_05408", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6090f63c611a48be6af45decc29eedb18239d475"} +{"id": "train_05409", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "reproducibility", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f0c628a7240fb99ae81c8740fe4eb1910e93c14"} +{"id": "train_05410", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5935622e0f812762e75a52b4d273f52e862938a"} +{"id": "train_05411", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36629ec6d2c6a0d0ba9f779bebbea014c8b1061c"} +{"id": "train_05412", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "governance", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d97d1e3013199658f81e454395d3c700ab8b0d4c"} +{"id": "train_05413", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54d061d3a51e8216f9de90628926f8151e26a594"} +{"id": "train_05414", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89069e4cf2d69b3e4d050b6c43f799d6a075032c"} +{"id": "train_05415", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "tooling", "documentation", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40e05c60338aca11ced32fe1452d25fa71b1b69f"} +{"id": "train_05416", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b264c661a517c1635a375d608ef0b61e8e56ef4f"} +{"id": "train_05417", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74d2477c207feaea0d5c6d2d647fe661fbef3261"} +{"id": "train_05418", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "091641a2f464359ce15e22aad700510ace19feed"} +{"id": "train_05419", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfc3f20585ebf4da1cd779ae83594351a19f3196"} +{"id": "train_05420", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce28d64eea86d6aa01de56454b3da47aada1d3d6"} +{"id": "train_05421", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31d961f4d27ccfed4a2b193f59f2747d1dbe8216"} +{"id": "train_05422", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c61c1a861d459c2bc1aa727a2cb16dbf2bd2eb44"} +{"id": "train_05423", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "reproducibility", "documentation", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "143e51ce16d181f5468d3a807649b0100cd64855"} +{"id": "train_05424", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "tooling", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a4e60b47a1636730209c95c17e6c51ac4807c60"} +{"id": "train_05425", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea98c5006cda634355c13efc11c97c755c1fedee"} +{"id": "train_05426", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "ci_integration", "tooling", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdfeb5c29f2416438e1f5fcfbb779a7c9ec84fb2"} +{"id": "train_05427", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tooling", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fe23dc23d8ad889fd610144ab9bb7086b9e8d97"} +{"id": "train_05428", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53dbb932fcff126220cd972bc3d4e5026bf25e6b"} +{"id": "train_05429", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "governance", "security_gates", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "139f5bff4a5875e47d8261d8b9ad22eb7b557e98"} +{"id": "train_05430", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "auditability", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53d231d47c57d29cec5d43339cc59b8688b840ca"} +{"id": "train_05431", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcd37e5ea1cfe178540453b6db8b6deff36b291b"} +{"id": "train_05432", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fd637716819d60ef8723ab7ac246b310fa13595"} +{"id": "train_05433", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "governance", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aaeff21e7f34ac625c7b036eb2107d6b1399c69b"} +{"id": "train_05434", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f193ba1c4ba36193cc932410cf95751ebda7baad"} +{"id": "train_05435", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87ad400fb436cce7274384967d6e1714d4b9c339"} +{"id": "train_05436", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tooling", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb46b44a5bdee05a218ce561055af994be992d53"} +{"id": "train_05437", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a98920fc5086dd407d52dbb629016804dd767d74"} +{"id": "train_05438", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bac47152894583a8a9dd7c9b91a510fb621ea5fa"} +{"id": "train_05439", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6744f99415ffbc33be867e82619536ec3e62bc51"} +{"id": "train_05440", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "auditability", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0af844fc9d8fbcd96418ffa693816155b488f164"} +{"id": "train_05441", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "evaluation_metrics", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f381f43d40fda4dbe816db7cb78a7707952ff585"} +{"id": "train_05442", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b06430161e2317c9009b2f5cf0189e803dd5caac"} +{"id": "train_05443", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35a079115756f9148bafd0ea54a181b48f8a1530"} +{"id": "train_05444", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "reproducibility", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a22725dfefb2272282ee5325e6648f38dffea9b1"} +{"id": "train_05445", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "907baf22d69990a88f6bb82094d6f017cb71e8f6"} +{"id": "train_05446", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "evaluation_metrics", "security_gates", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1155089bf70a72a09df0e3acb9748e71f3a3941"} +{"id": "train_05447", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0539173409c183fdf9e16c7fffd41f2eac133fd1"} +{"id": "train_05448", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a589956a6669f01665b19c70169b701fbdf66779"} +{"id": "train_05449", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "documentation", "governance", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8b4ca2fb4d6d67661cf27726e79b51cdce3e082"} +{"id": "train_05450", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "108f4e4abc1ba3116ef1c3f4e30c47f0a90646db"} +{"id": "train_05451", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f2bd183c4ca8925bd4315f05d72d87f9ca9cd1c"} +{"id": "train_05452", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b30019b198a402048bb7d949a3ded373acf4844e"} +{"id": "train_05453", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "governance", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f04b3160916a98ee9f09f7f08bda8645cfa022db"} +{"id": "train_05454", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "tooling", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c70d358af34231572a0e6af1af622efb0c8e38a6"} +{"id": "train_05455", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ba357600293d27c9a5c5f95551fb140b794f104"} +{"id": "train_05456", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d137a42824d07565c3ea287fed7a43b448dd044"} +{"id": "train_05457", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1bb897aadc1e9059c721e8a82884ede593d577e8"} +{"id": "train_05458", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "governance", "auditability", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2329a985df355ae3ac87328b3ec9b47bda6c2c04"} +{"id": "train_05459", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61c7cb644a16228bed60784ce424abebd1e87b15"} +{"id": "train_05460", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4698727a8e8cba4795af1b5491a398d059fdc87"} +{"id": "train_05461", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25c044c4b7412ae57af73a9bd74653ba918ef589"} +{"id": "train_05462", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "security_gates", "documentation", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "606e11caf555c311b80dd1bd22ec91d53d7f1011"} +{"id": "train_05463", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "documentation", "tooling", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37b77d0750f355970b2859b483ee1bcb5a52a77f"} +{"id": "train_05464", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6f4542753354d3cd7084858f4b007bb2dcf9fc7"} +{"id": "train_05465", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4665c42ff26b6988aaa18cbd70687ca6eb532b4d"} +{"id": "train_05466", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "governance", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08906fc1af6bfe02b620e1051db49bb0701062cf"} +{"id": "train_05467", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tooling", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec9ccbdde06c0f03ae7bb026153fa63a9c15e49e"} +{"id": "train_05468", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9374dd4b3a94492eecec75a00b5c481c6e45d6d"} +{"id": "train_05469", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "918f2ecf88bace9ecc9a904aa73afc6fc0e3c2da"} +{"id": "train_05470", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "documentation", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17df53066f6e3d7d278c9182cc2b108f2c8b3bc7"} +{"id": "train_05471", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b47eb1cb9c8b21a7fd92b61f32f8275b9599836f"} +{"id": "train_05472", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a1a3b6e2d04e75789b086f743a5003cd28c930b"} +{"id": "train_05473", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07932a706c9a3b84748ba2097da5261c25e715ee"} +{"id": "train_05474", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10da8a8e5153ac4cf8949d9c6762c8120854fa7d"} +{"id": "train_05475", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "512b8bd43a126e6758ac4c47fa6e67bc70f25eec"} +{"id": "train_05476", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "repo_scale_reasoning", "reproducibility", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a98a4a917cae83fd3280c5e4c35f464b6518b3c8"} +{"id": "train_05477", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6ec637b644ba1a86d7d71b27c7287583d31538c"} +{"id": "train_05478", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tests_are_truth", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1283313eef4a13da415618c733be68f3439799f"} +{"id": "train_05479", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89069e4cf2d69b3e4d050b6c43f799d6a075032c"} +{"id": "train_05480", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "repo_scale_reasoning", "governance", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3709b4cffec0c76f1cac074cf25d2f2683a023f2"} +{"id": "train_05481", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c71d8bac33cd0ac52721b6ecf02d02967f8f4fd5"} +{"id": "train_05482", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "reproducibility", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b9efda42661b6fda9df58abddef68e7577686b2"} +{"id": "train_05483", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "evaluation_metrics", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3b8f4eb0aee6de5be894fd9771d041cd9375cda"} +{"id": "train_05484", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "documentation", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b099243d1bf7df9e95b17776aed572736b66a7a4"} +{"id": "train_05485", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "documentation", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9565d8978670fd2cda44bbba87fd9481368134af"} +{"id": "train_05486", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab7800b3fc708cc87011c82f2621d707a822159f"} +{"id": "train_05487", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98f92aec0523a1a55fc0983937841ca72ec35a3c"} +{"id": "train_05488", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "ci_integration", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f21ae623da4dad7506ae0da975e7b676503961de"} +{"id": "train_05489", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c597533f7908ee0e338913d2c6ea52dedb85e68a"} +{"id": "train_05490", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b270af7ef6e1346eea26689289ec759900e10f60"} +{"id": "train_05491", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "992f738a2a8083a861d529d9d69a82abaff444b1"} +{"id": "train_05492", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "cost_latency_tradeoffs", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a6de7df2bbf2663b028d43b329ddfc2f0212976"} +{"id": "train_05493", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed07d2af8b41133b143c0c53601b417bf3d54d96"} +{"id": "train_05494", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "786a4a0752bcce71f89651e2fc63817698b58896"} +{"id": "train_05495", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa89f599f2cbcdf0226cf81de46ddd76a5ae7b4f"} +{"id": "train_05496", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "documentation", "tooling", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39ddc80008f3b79700b642d8e5338dc300200013"} +{"id": "train_05497", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "925c5bb6ce92b69909db0ff6b1aeb7157b54053f"} +{"id": "train_05498", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "governance", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa04dfeab25da1091a983da242a4e39bca0b7937"} +{"id": "train_05499", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3110e3da9bda398463ffc64cfdad8a55351bc93d"} +{"id": "train_05500", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d866789c9298263fde0734864846bd354cf0f707"} +{"id": "train_05501", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tooling", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f921a2dea01cef105423d25497b39436b38eb1d2"} +{"id": "train_05502", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04033ed402e1175df1b643ca81927d22931c6749"} +{"id": "train_05503", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff05e91a734e30e2d1d79c2ab3b5c73886cc2bd3"} +{"id": "train_05504", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5217aec653af3a41e3cd0b26e2eddc17bb0d76e"} +{"id": "train_05505", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0125a48d7a92afb403a7d162ea2df7df73c29f9c"} +{"id": "train_05506", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "documentation", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66d69fed7220905aeb7545ee1b05e6794588c9f8"} +{"id": "train_05507", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "ci_integration", "governance", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23fd7042780392ebbc1584ab50a5ea1d5bb1c545"} +{"id": "train_05508", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ead3831d6643d2c1fcaa1f84e57da28f7c7e50d8"} +{"id": "train_05509", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad1bc191f2ad801cff55ddc09db44e0e44c40a8d"} +{"id": "train_05510", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bff923dd4f65887cf72856ba328e9b2df3e3af99"} +{"id": "train_05511", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c0cb02d2979ef213dd4ad360239e4eb79801b74"} +{"id": "train_05512", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66ed419da1c55e0ad18cc9a61e68626991619c37"} +{"id": "train_05513", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d94f42db64babdf1d48e98fba0da10482601895"} +{"id": "train_05514", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4d49c794ee288ce5fc92f1aea59ed6814f5e232"} +{"id": "train_05515", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "482d9cc967a5428bcf39dad8c71a8ae955682328"} +{"id": "train_05516", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fa89533996246f7e6aad57117025622065e111b"} +{"id": "train_05517", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a203f7ff8176ffd4bfbaa3a0a1dc19cbc019b92"} +{"id": "train_05518", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f09b499fa2456b41b7696a21d520caba0d1e363"} +{"id": "train_05519", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "evaluation_metrics", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89e3a31e89aac099e8d1a77b338028415e31771f"} +{"id": "train_05520", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3e78693f1280e25526350aaa742ba386e258583"} +{"id": "train_05521", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "443ec1c4e0fac27110083e4162747d7c4e51aa97"} +{"id": "train_05522", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "431bc81728b4f5f96f8b7fd1ac4a855cbf79d3c2"} +{"id": "train_05523", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd1b1127d241e46b6274f42769b23797edcbed3d"} +{"id": "train_05524", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52847a43b9689b2ea74e3df7e88b43aa74595711"} +{"id": "train_05525", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5afa86e27d0699ec1927dd6dc048eb8de77c557b"} +{"id": "train_05526", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "207dca41468ce9ad1d2da025484524decf648aa0"} +{"id": "train_05527", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88f0d5faa706d0daa36419c871db959d5674cb65"} +{"id": "train_05528", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tooling", "auditability", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d056ea3e21c528c405fc3161203d20642c38705a"} +{"id": "train_05529", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "repo_scale_reasoning", "reproducibility", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "abb3da8d765df5f6b947c70baef2d7bdae47f205"} +{"id": "train_05530", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ac559a00a0df6522d96471c8d2f84068cb2c166"} +{"id": "train_05531", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b76a148e23237b63d2e9d6088429c9a20b7068c8"} +{"id": "train_05532", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "repo_scale_reasoning", "governance", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "617d91b037191add1855ce9f42b1a746f1cdb2e1"} +{"id": "train_05533", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b9b4e61e80f10413c88d26592b9fbfe50dc9092"} +{"id": "train_05534", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc13d560a57145418916f5014c90631f2caee234"} +{"id": "train_05535", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tests_are_truth", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de7990b9141bae75582d13fe01d826c674b79114"} +{"id": "train_05536", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "ci_integration", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42bd8db05e17f9e1af4aa2d8a538ea9b8e9f03f3"} +{"id": "train_05537", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "documentation", "security_gates", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9338258b568e59252aa092a02220ec72c5642476"} +{"id": "train_05538", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "170fc47b024bddacbf4b9a92807ef35f179ea613"} +{"id": "train_05539", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "reproducibility", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5c04876c3a0da32658605e0f0244a311d736d06"} +{"id": "train_05540", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd93d1aad9990a6583ad9c6fa2e3957d024b9d16"} +{"id": "train_05541", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92a77c14d72843a238787ba51c64eb059684f951"} +{"id": "train_05542", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1832377fc41cceb39f7e9f4a856914d9321b6914"} +{"id": "train_05543", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "security_gates", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c48f8250326cb8ddf4af95491be28a0bbcad33a8"} +{"id": "train_05544", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "documentation", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ced8645d458f12c1cbdd27f87896fc9ff1da887"} +{"id": "train_05545", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ffd5a1686bfa26401ec05ea3aa21b56f96bde27"} +{"id": "train_05546", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "security_gates", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb2ff79bf0d5c2b1a0d2db702d8fde183fdae6ed"} +{"id": "train_05547", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c66af4213edacb054705ec6bdb118cb0b66c2d0"} +{"id": "train_05548", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43907aeabd79a6856396c895e701606b70d61638"} +{"id": "train_05549", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8277ad08cc8b51c787ece9d5b8cd375c1d953a1a"} +{"id": "train_05550", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a265d862bb6f00b5e8e0d5208c6087785b0e12f"} +{"id": "train_05551", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "security_gates", "documentation", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecf845c0360a6107329476d53405fd9c16c16f67"} +{"id": "train_05552", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6a70545ced29ad791d68bb3f3edd229a94489af"} +{"id": "train_05553", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c48f8250326cb8ddf4af95491be28a0bbcad33a8"} +{"id": "train_05554", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "documentation", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b9efda42661b6fda9df58abddef68e7577686b2"} +{"id": "train_05555", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de1c24b23069603c71692acc61572bc3c78de901"} +{"id": "train_05556", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "reproducibility", "auditability", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13159a0bfdac9b0e0dd89b1d8f16bbc26649c7d1"} +{"id": "train_05557", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "governance", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7508eae05c649e6ef8f0c3b9a6154996edfb40ad"} +{"id": "train_05558", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "reproducibility", "governance", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d516bfa2c1db9fb66cdc193dc9ccbf34d3c25825"} +{"id": "train_05559", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dddf17491d12b3c9c264ca45c0e560233755d94"} +{"id": "train_05560", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b8f1cffeaf583e7a1225e0c591e3979eb432146"} +{"id": "train_05561", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afac449107699a759b9b3277b9627e7843d3c28a"} +{"id": "train_05562", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6197b5dd8d1c7d40c94c1ea15d3a24fc125d987e"} +{"id": "train_05563", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4797eba72ffdd1dc33939fdcb3172827708814db"} +{"id": "train_05564", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "governance", "documentation", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ba04e21cdfc2105b21dc5411de05e458c2a863e"} +{"id": "train_05565", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb48a5a771e86f4e725b495f468fae4cfbfa1e7d"} +{"id": "train_05566", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad07fcdc4d24df4747c917f443b6938aa87467fd"} +{"id": "train_05567", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "documentation", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afd88ed6cec59649821f895b05ecfbd1da6da449"} +{"id": "train_05568", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39f29b3d1fe9697d4e78dbeaefbf457878b4c023"} +{"id": "train_05569", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "security_gates", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "843ea2b61043a0d10b6cfe8ba2048a787fe3b0fe"} +{"id": "train_05570", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1e3bd1417cdec4eec36173063b6f1cd0161f159"} +{"id": "train_05571", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2971397baac89d3d00cbc258453f08cee33f7afd"} +{"id": "train_05572", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdf83212b2ada54c7d73fbffdd81f6ca6df95d40"} +{"id": "train_05573", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8866f9a621e472ce624a7f0400f7b577a190da91"} +{"id": "train_05574", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "governance", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c82ed0b0b6d355ce7c9f5ef42c1dfa5db096531b"} +{"id": "train_05575", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "governance", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b9b60cfc59e849080ccfd54e4ae48e07a72d889"} +{"id": "train_05576", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3389e8268935757a265f3beb7caf80e18c0085a2"} +{"id": "train_05577", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "security_gates", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "102737d0d26a3a6ff789059a31299546a73f4cf0"} +{"id": "train_05578", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "tooling", "reproducibility", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b27058a1ed1d7e90645fb52358583b64c1c7e8b3"} +{"id": "train_05579", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8346334251620af16454d8a44835ed31d2d31b52"} +{"id": "train_05580", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "febcea60a10345a7668d4f2f4b7181b77d9f167e"} +{"id": "train_05581", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58d274f1057d70a306b16d7b635078281a0f2bba"} +{"id": "train_05582", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "documentation", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d420a7b05128d81da5dfa248ac6dcfae0bf7a308"} +{"id": "train_05583", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d83571eeaf64981ac38c62742db228e2e4cad4f"} +{"id": "train_05584", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "evaluation_metrics", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66d3972c2210bd8e67b66276bd12d2f7b78607f6"} +{"id": "train_05585", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "repo_scale_reasoning", "reproducibility", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c455f28ea13c5abdb49ad9c79e53cffcc00dcd5"} +{"id": "train_05586", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8349827930755ee3a70e1320fc715805b329c60e"} +{"id": "train_05587", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "governance", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ba5631023ae1afbc26aefdafc2719c394befadd"} +{"id": "train_05588", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff167e2dd89774f30df63247c7a06cd505951765"} +{"id": "train_05589", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17629894c60819948693a10fc1f3c8f14f99ac9a"} +{"id": "train_05590", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37ccc1cf4debd1761b8f8d298373982a0b3bea69"} +{"id": "train_05591", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "reproducibility", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d39d2fde3133e3acb5f467fcbe69464817485100"} +{"id": "train_05592", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "security_gates", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c982d3fc7f1f618ca19c7b0d2e726e58ffa5d59"} +{"id": "train_05593", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbb19da14ceca463fa83395f7d3a5679baa5e8d2"} +{"id": "train_05594", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d0e6dcae8e92872ff86bcf87ee5c011cbf3a372"} +{"id": "train_05595", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23a1a42b97b53afad135f5e65529298b0bc9c9fd"} +{"id": "train_05596", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "documentation", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89949ad9187b7b0b63d898af5d26c404adce0f22"} +{"id": "train_05597", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "governance", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfc13c69d7e58154df725384a7b1d4f3c22b415d"} +{"id": "train_05598", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b20d3ff7b7f9e4b029fce0de1b3ff7d3d9aea10"} +{"id": "train_05599", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "documentation", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e709e9e4a3830e7da345448da96b4d64dc539a6b"} +{"id": "train_05600", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca93159e710d65756f7b1345109941be0629dad2"} +{"id": "train_05601", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09d90524b6f6ce82e48abebe5550f5dbf5ef2123"} +{"id": "train_05602", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "governance", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bbeead378deeb68bc7282336fff0d0b011907fe"} +{"id": "train_05603", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "869d61c20b0b80d59282aa763ec06e2a07e8123c"} +{"id": "train_05604", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e8fb82d2b50fd6411d2052e27f4f5f4cd4a2620"} +{"id": "train_05605", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4aa2a38ee97d1b1519e44f2004503d1a588f5e73"} +{"id": "train_05606", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "security_gates", "auditability", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2436c48dfbde4d84ee2f1048ed52f2d9279016b"} +{"id": "train_05607", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "952541e8906b4eaa3bd83e93cf60ce4d43751a08"} +{"id": "train_05608", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "ci_integration", "tooling", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5007e4d8d36b22e40260f5e55ac8d2d685514d6c"} +{"id": "train_05609", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70826dd52247054967a20032a1169505a6019b19"} +{"id": "train_05610", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "reproducibility", "documentation", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d39faa371d5a7b12b4cd6de27b1debf8c05191e1"} +{"id": "train_05611", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2777281a6939c008f19082ee5d4d657b6b8b3028"} +{"id": "train_05612", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9804f74cb2db74e5e3cbba628bc3ecf03f9f9a8"} +{"id": "train_05613", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ab162663d2b7af924c71f82d15d66bc10c7e743"} +{"id": "train_05614", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4494622300d4cef59fa5c01d91982f3861b1360c"} +{"id": "train_05615", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "376e153833383bb8234553307a0be99de68499c4"} +{"id": "train_05616", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "32ddfbe43792e6511301b6089a558105b9606a8e"} +{"id": "train_05617", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99d6acf94a77c30ef96d6c6caac297908f7f6208"} +{"id": "train_05618", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "074205342f77336b6fe5f0db334551ade8cac605"} +{"id": "train_05619", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "tooling", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ce9a5701c834cb13f93a5d03aeaae8bb6a8c3b8"} +{"id": "train_05620", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94b13a16538b57baf55c0634176a9013dc5edbfe"} +{"id": "train_05621", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d57c8c65901ec53d8e0ed53ffa764ebd32cc8513"} +{"id": "train_05622", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fbe83735bac342372d7aa79eee94357f38790f9f"} +{"id": "train_05623", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "ci_integration", "auditability", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8aeada899da5817f73794b55aeecdc86e849604"} +{"id": "train_05624", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "auditability", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a697976e2f02f6100bec23286d2bfb601efebe65"} +{"id": "train_05625", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9780a9794513b467e02aa370bd65509609c6cbd7"} +{"id": "train_05626", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "787a4f2e5ef65c6ec6591dd56f9e3eed1288295e"} +{"id": "train_05627", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "governance", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "faab8848fade1809ab4d4bcbe8b7568d2e63eb29"} +{"id": "train_05628", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "auditability", "security_gates", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c09a8cf389491c562270d3eef0d335e86ad9143"} +{"id": "train_05629", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d89fd6342392d5e9d9853618ea25c4bfe695139b"} +{"id": "train_05630", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e743e5e89f4672b4fc2ef1fed93615e3000d4461"} +{"id": "train_05631", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "ci_integration", "auditability", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a895e8916c1e99264903f5c72aafb94c2bb3aab2"} +{"id": "train_05632", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62c6872f057c38a654032d9894a45c83bdeae1e2"} +{"id": "train_05633", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e37014ec026884ffd003e78a8e81a549636a8b72"} +{"id": "train_05634", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4af42fe97d6bbc7d8afb6e37fdda93b098536577"} +{"id": "train_05635", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a39dc39b21cdbfb15e6ef27606049ea399ab1f2"} +{"id": "train_05636", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "492ba6d8500e00f03fc3dd08024f24f7423abc9a"} +{"id": "train_05637", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2169545f652da9c91dae2cfe11d50ee0fba66ac0"} +{"id": "train_05638", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5595c1cb220ee5c6c2cfde50a20df11af6ff7726"} +{"id": "train_05639", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d804f9abe36074cda88a1465141af707496f524"} +{"id": "train_05640", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac2a7c1cc3fd180706a8ae7e98ff243d1daebe24"} +{"id": "train_05641", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "tests_are_truth", "ci_integration", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33e8ecb96ba4c1b408a15928cd2eb54449c39891"} +{"id": "train_05642", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e3fae20c30ebd2c974c6b00365c6ce4fbd882c4"} +{"id": "train_05643", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "auditability", "security_gates", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11e46ad6f2af649fac7a42bce22bba6ab882af9d"} +{"id": "train_05644", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "security_gates", "auditability", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b3b39d29d50323687f31cdd02ba47dc558580b9"} +{"id": "train_05645", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tooling", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48d20b38d81f3c61104614effc787756300a6708"} +{"id": "train_05646", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ea34a72c81cab07160542f0b862224dc5113cfa"} +{"id": "train_05647", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "349ced3791f5d34bfd762d74e650006c21a4a2cd"} +{"id": "train_05648", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ffe3779ba460c3eff527f6dab38f241d97ab537"} +{"id": "train_05649", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdde8b5c36cb59ae1f692edadb3dd9a16b977780"} +{"id": "train_05650", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47a45c3bbe827362d434b8f7b20406e1e74d2737"} +{"id": "train_05651", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "auditability", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "643c4ece41aa36fbad01085fc6e30ca4ae7e1711"} +{"id": "train_05652", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c222c36d559b0e5de74d827659e01d71bea58ec1"} +{"id": "train_05653", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "reproducibility", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47163cc18a17d7ae41ae866cee9e469c7f7c07a4"} +{"id": "train_05654", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "repo_scale_reasoning", "documentation", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2555610479d82f75196908f0d1eab2df0d4ed0f7"} +{"id": "train_05655", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "ci_integration", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43d235f0578dcc347e1a9a73ae6d3e89d9e1cf2d"} +{"id": "train_05656", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "documentation", "tooling", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a924fb2dc1e5603a44f8e3000b2401cd4fb76f94"} +{"id": "train_05657", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7509a4795ad57bee202c645de78f05e1cb790b01"} +{"id": "train_05658", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb8d0c525914ec1ac5294df12b7a3ff701b4196f"} +{"id": "train_05659", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "security_gates", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e50eb6a619704b7ee6fe5037ea59a0fce37c3d9"} +{"id": "train_05660", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "security_gates", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0ae334f216fd2d603ebbac8d83d78d8689cb942"} +{"id": "train_05661", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d84587e58f206bb969094dce3afed5bd5acdfe9"} +{"id": "train_05662", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccbb92ac8b8056c09f743565d8eb7dceb74bd0df"} +{"id": "train_05663", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "618f97f315e3fc62601c574a44e3111012f37b3a"} +{"id": "train_05664", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4913f029509d6d88cc78b1f7dc260ab7444aed21"} +{"id": "train_05665", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tooling", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "743e2bf74b7e14981e02382fb135c09c48db1a35"} +{"id": "train_05666", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e81e058818e9e0021192907edf9416afa9685512"} +{"id": "train_05667", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2405072b41ec6c25e732f4bb7adf45a8c19c0a92"} +{"id": "train_05668", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "documentation", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "918f2ecf88bace9ecc9a904aa73afc6fc0e3c2da"} +{"id": "train_05669", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dcd11194239811c92896469931eef0ce503c501"} +{"id": "train_05670", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "014e76041cd70ab99dc76bff03420108aaf2ee8a"} +{"id": "train_05671", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "reproducibility", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f086ca00037f29734852e94c150c7cd20ac723f0"} +{"id": "train_05672", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "auditability", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8367d7dba0b969c53c553f1a78c9bbcd1802678b"} +{"id": "train_05673", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b6cf117f3723a077de7855e5a35cb38985689f1"} +{"id": "train_05674", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "repo_scale_reasoning", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bb00640aa59420b8bb9d34bfee8aaa5cf172d01"} +{"id": "train_05675", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9571e5c91be2d7ab32551c63fd251e48639e10d"} +{"id": "train_05676", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "401f5bf072879d7585db2b9855da360082772006"} +{"id": "train_05677", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fb18aa89a4958ce92a72c31532176240fb38b4e"} +{"id": "train_05678", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1832377fc41cceb39f7e9f4a856914d9321b6914"} +{"id": "train_05679", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "144ed43f4dfb22253303006022df3144c5c05a77"} +{"id": "train_05680", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30e4a821a1e28ecac122d37d0b16f390b2b6499f"} +{"id": "train_05681", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "reproducibility", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "787e870cae58da2bca0cb25c6be09c8ac506be81"} +{"id": "train_05682", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "ci_integration", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7632bca8831d1b73b293b27090ab30705ff75683"} +{"id": "train_05683", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8fa802a26140e973ed6ef33f8caab965deda83ac"} +{"id": "train_05684", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec4080310099c11c21c2e3163b2246293362921e"} +{"id": "train_05685", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "security_gates", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8a3b92f322f34bf1cbdc818e31b905d0a4687c7"} +{"id": "train_05686", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "auditability", "security_gates", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80026ba07acf720ba9937237aac6c4615d368d58"} +{"id": "train_05687", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20eea22dfa1722c29d44ba009fe90c0f2f87d47d"} +{"id": "train_05688", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85faf0e358f745f27f142c7d2aaaf4cb03af7210"} +{"id": "train_05689", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af46df9a16004c19fe3deb42329d62b19c84ab8f"} +{"id": "train_05690", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "reproducibility", "auditability", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "555c01fb5e6e602fe7c3ac5c168257517ed3f3c8"} +{"id": "train_05691", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a8edc969e135cbbe381c1516598c9add456b07c"} +{"id": "train_05692", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77461fd72bb73d847f12599e19be50f4a6da4444"} +{"id": "train_05693", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82e35e4d43cb17d0d973c363d9cc55202a50b963"} +{"id": "train_05694", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "704af11ea185e03c57da56cc36fdd688fb01fe74"} +{"id": "train_05695", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0eed7df232b1e461316c94530635a0ed6986dde5"} +{"id": "train_05696", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "evaluation_metrics", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd8dadb4f976e16463fcfe4bae9f561ef5c204cd"} +{"id": "train_05697", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "security_gates", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2b3db1796d1b8aec8d080a1e4e32fedffec4a25"} +{"id": "train_05698", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d39e2cb1a32b3229c877982178378cc7c1a011eb"} +{"id": "train_05699", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "606e11caf555c311b80dd1bd22ec91d53d7f1011"} +{"id": "train_05700", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ff00073f58c8a5c62b1ded253f0f73bf9c22900"} +{"id": "train_05701", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0071a466c65c27287f837cd384f504d836564603"} +{"id": "train_05702", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69a09690fcb9f006a5938eadb051cc9444d7c6ff"} +{"id": "train_05703", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "evaluation_metrics", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f192f8f671a71a794e78cc7a6984b17216ee713"} +{"id": "train_05704", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "auditability", "tooling", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3908fd3e2d068d8b7416acbad8fbb808641585fa"} +{"id": "train_05705", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f9641c4153b30d314fc22ffd8f653ea9e8a01da"} +{"id": "train_05706", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "governance", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83a67c6720eb103c2637020132a7990385082d5c"} +{"id": "train_05707", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tooling", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8ba36120577e346a20239cfad8fdd4cddd08937"} +{"id": "train_05708", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00804b1680af52e0e72825eae2b660be1d237adf"} +{"id": "train_05709", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29cfb9ca9c1fdf5fbfdd361bf31caabc269f55ea"} +{"id": "train_05710", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58498ea86b528aeef4ff81f76e19ca1e76eee4f1"} +{"id": "train_05711", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "492c12e8a64e2de8ea4c56ced652485430c01f9f"} +{"id": "train_05712", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "939ef8fde879a86440959e863f743b4e031cc4a9"} +{"id": "train_05713", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d42c926ae32f8e9eeee3d0366e9a4a8308acb60f"} +{"id": "train_05714", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "reproducibility", "tooling", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d18408230c07b78df116c197869959e443906578"} +{"id": "train_05715", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "evaluation_metrics", "ci_integration", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d929118167b483ada6c22e092112bc310b890d8d"} +{"id": "train_05716", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2f3bbd8910693a634222998a553a655365fb14b"} +{"id": "train_05717", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b126a8d0418210478d58df6b8c2add0c0271aae3"} +{"id": "train_05718", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffa4b540abb8431a5bb02aa67520c96aa0051df4"} +{"id": "train_05719", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45b4708adb85855e3c2a986b2ee6537e879b1330"} +{"id": "train_05720", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4dc2ff91481d76d165df2956108f88bf3bbad8f7"} +{"id": "train_05721", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "governance", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffcdbd2d215bb0aa61cb2a52fb30386ceb4138f3"} +{"id": "train_05722", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "reproducibility", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b5844dcd942422260ce1871bd207886a20b998f"} +{"id": "train_05723", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "governance", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce3da8c12e160fbd979ed63925ca94a0057ec627"} +{"id": "train_05724", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "auditability", "tooling", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6df90cb7cd7f1fbb39f563ea78414e16c836f3fb"} +{"id": "train_05725", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "134bde94580531c1e8cab356890a9d916c4b42d3"} +{"id": "train_05726", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5e54f917d5e06c24e5903e6ad5332b19804066f"} +{"id": "train_05727", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "evaluation_metrics", "documentation", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed312d31d2955b82a86d2cda769a08c499772bfd"} +{"id": "train_05728", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50292ca484df8f4bfadd5a637d136f706c22d3ff"} +{"id": "train_05729", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "security_gates", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99542cca70308c9466dc68c7c2a2165295f93cef"} +{"id": "train_05730", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ca419b554a3e743939777d914aaa1c4449d7143"} +{"id": "train_05731", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecae17113dc7cb4c510afd8036be942d6c964721"} +{"id": "train_05732", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "725f94befe7296127ac2cf81472451faa6f07212"} +{"id": "train_05733", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "084bd250e639e3e9ee8714794cd3886dd3426a0c"} +{"id": "train_05734", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac92294291de53a7f6b4909dfc03e63fa34bd755"} +{"id": "train_05735", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "auditability", "security_gates", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81585c228021d39a6236ac567581dab83f9f8a8d"} +{"id": "train_05736", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "evaluation_metrics", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6202d627269b367468269a7b716e38f62e27fdf3"} +{"id": "train_05737", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "114fdbf067691158250a6e235989459629f518d9"} +{"id": "train_05738", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "reproducibility", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69191b4da1804360dc15ae438871a03f906d6ca9"} +{"id": "train_05739", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "auditability", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36a54b4db525c8afcd2ff5181845040f072c52ca"} +{"id": "train_05740", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53223f362ae735ee800726f98b78c524e7b1bc30"} +{"id": "train_05741", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b9f7253288f843a14f66e9cda8fe154db898b43"} +{"id": "train_05742", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "documentation", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e02c712e26524ffc73fab277f13c283d3e00664d"} +{"id": "train_05743", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cafb000d1a43e50ad8194ff0d63931ced50aac36"} +{"id": "train_05744", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd1411b5786e21d6d105c366a70a43641303bcc0"} +{"id": "train_05745", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec4239d7dd33a08460bae772fa6e8fdd50d1125e"} +{"id": "train_05746", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e403c6c1c136226cb99fc37d6d148d215a648e12"} +{"id": "train_05747", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "auditability", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9acac950b622f8ade24a8b56fe52ef55c3a153c6"} +{"id": "train_05748", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dce10a55364fc09a0f2e042258b17e25a5b8a2e3"} +{"id": "train_05749", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "tooling", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2f964b3dd063776c1c75710c84f2e8f68e73486"} +{"id": "train_05750", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f56bed9b7ce6e49092da83a6570d63af613bd8b8"} +{"id": "train_05751", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "644f43bfa5f169b775a6d7e565a63bbaeaec811e"} +{"id": "train_05752", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "governance", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93f189aac43befb83356615ed1a4a5957fefd67f"} +{"id": "train_05753", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a397bbd66c73b683f25e10c49e6af49741074ceb"} +{"id": "train_05754", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "documentation", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4061492f1bda2f7f882afe0047e4c6bedca1ab19"} +{"id": "train_05755", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "security_gates", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c702830f27a0f4b564406364989b9fb9e201ae83"} +{"id": "train_05756", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "security_gates", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f6b99c1e81fb65940d76d693d4be0ce9bf42836"} +{"id": "train_05757", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "security_gates", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ce17238cb4e00d87940a904e4d5646f0b416a25"} +{"id": "train_05758", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "baa141b8b9656c4ab0504e807840b944812437c2"} +{"id": "train_05759", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb27f496e101c0e14cdb9884b65f95ca2e489060"} +{"id": "train_05760", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f2be444f3bfcf11185897beaad796213672a984"} +{"id": "train_05761", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c145811d6bfb5959ffa12dc9fa1aac516807ca0"} +{"id": "train_05762", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b0fd8a4558de4f90a96e5c2070124656f970910"} +{"id": "train_05763", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d67064746ce0e311d7c20c5452ee8c3ee6fdded7"} +{"id": "train_05764", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "documentation", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c57871db5e158d3dbfe86cbc976b71a7fcfb2079"} +{"id": "train_05765", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e871e49d52e7e1ec1c0e3456f5a20f1bdb50b02e"} +{"id": "train_05766", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tests_are_truth", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3af11f24d4d2e9e6e963d078afb74789d84ec61b"} +{"id": "train_05767", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bf35161b06acf436ddd66e2e199c95350357002"} +{"id": "train_05768", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "446d52e40a60968652f70686e919dec41db0e427"} +{"id": "train_05769", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67674507638e4253ade2509ad90b4e0614f90f1f"} +{"id": "train_05770", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8b4c3abf185e477937f6cdb983e116b60971a2b"} +{"id": "train_05771", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df73bdc1c252852e75ce13d0a88dc290902ffc75"} +{"id": "train_05772", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "security_gates", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7de1d117a9c9b55e37bb36360572d5d0fee614a"} +{"id": "train_05773", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ff301c26872433601a84af20680901e667070cc"} +{"id": "train_05774", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc0c2ee21295ca345dd0ee558e75d8fd41f9ca78"} +{"id": "train_05775", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d9517708950cb874f507937c737418eaa167a4b"} +{"id": "train_05776", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea2bd6d616dce9d34c28d81b3a7f6edf2696a56e"} +{"id": "train_05777", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "tooling", "documentation", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83f60acd9fb32fc158de1f68ad70250f69ea2f8c"} +{"id": "train_05778", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d66991a24ebe8d5a42a0e2b5232c1e314331012c"} +{"id": "train_05779", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "auditability", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ce6641fe892c3f9b874c5e34151eaff4861b49a"} +{"id": "train_05780", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ef5e8ea936da4255d43d873cc345848a86266b1"} +{"id": "train_05781", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1a6d43be1e78ceff0d95bb4f92dec2eb92138a8"} +{"id": "train_05782", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05d5c9538790a889f6185b61f4d6f0111481d7c3"} +{"id": "train_05783", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f40db5a87779abe48be3c4a4676c91482da32f90"} +{"id": "train_05784", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e96ac32fde18709fe30eb74f39a87c68f6de4aa"} +{"id": "train_05785", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c629afe13c55743c427219b82dd3fe0a040e460a"} +{"id": "train_05786", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "tooling", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a553d83a955a8a69c4b918acf2815e3085275d10"} +{"id": "train_05787", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22f42f8c834e48a87b24b981d625f0b5e1f82e98"} +{"id": "train_05788", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "auditability", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "735db2e5f6c2c3163491b5bc41dc3b7e0b91e178"} +{"id": "train_05789", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "governance", "auditability", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e20c64c00224e99cd0934b5fbb3d0f87cdbbf9ad"} +{"id": "train_05790", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "acf1642116e1b04dc901da3b8592774424f3a2c4"} +{"id": "train_05791", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "reproducibility", "governance", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0592571c7cd1591ab65fc8b4c4ef74be48d4df51"} +{"id": "train_05792", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06fbbe411dc38b93aa158251221c60802689c7c6"} +{"id": "train_05793", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c43f008023ac5502f461ab1c70821ecbf5eca79a"} +{"id": "train_05794", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "evaluation_metrics", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d58a125cedfaee93464607b768803c96aeba1e2c"} +{"id": "train_05795", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef96c06b568f12dc72a6ca5eec4e81085e595f80"} +{"id": "train_05796", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a8c7b86f387f9711625e72d125e02ccbac34d34"} +{"id": "train_05797", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "830a9fd51f5b725eba8bce7cb84d79cc33201fd9"} +{"id": "train_05798", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b8a68184df6741f884e413fc7cac30635e64351"} +{"id": "train_05799", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "documentation", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28bf92026ad43679236ea301f97c41dd27a9cd0c"} +{"id": "train_05800", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "747b02c225e44dd84f38ad090f7d8927025bc671"} +{"id": "train_05801", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5e3e204c8d1ca25bcb7cbf8f02d5eec46c70999"} +{"id": "train_05802", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7db73e10c6ae2686d9888f0f2ce7e078d3b5485"} +{"id": "train_05803", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15c103ba376755d289d36fc591c533f4dcc1bc49"} +{"id": "train_05804", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3424a71bfdd07d575f26e86218a3f477209a94f"} +{"id": "train_05805", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64aef519872aa1b1aff39bce87373c39836bc4e1"} +{"id": "train_05806", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7654439e8bd6e3280856e518cbecf4a553e81005"} +{"id": "train_05807", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "governance", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20aad067b934778bd91c458f317a0688e3c31f67"} +{"id": "train_05808", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37ac29b012721f565722cd3b520d428b096163b0"} +{"id": "train_05809", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "430738217c7d153b5374af7f11e6011ef4a97b5a"} +{"id": "train_05810", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c83453c3d0f32664877daa616fb15109cea98205"} +{"id": "train_05811", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8deae43905bd3e9d2a1d4b216c5fef0b751b562f"} +{"id": "train_05812", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76a46615dcc12079f79e617b651d1a619c0611a4"} +{"id": "train_05813", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64f3019544736c5dbafa95c4510102e00fd87732"} +{"id": "train_05814", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5d0435432a99685a44bc8fce37180ba7f456a71"} +{"id": "train_05815", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fa156ca94506031b6b36b176935cc921beaa006"} +{"id": "train_05816", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tooling", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ce86311b13b4cecd4184dde0374ce3bcf1e50ff"} +{"id": "train_05817", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f6881c253590cf01c3a73a165ef0b56be72f6be"} +{"id": "train_05818", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "388d2676ca228b3ae4d8db42fff14746b7291c0d"} +{"id": "train_05819", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tooling", "security_gates", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d74a9e311c01555c929218fcc48e9553668b09a"} +{"id": "train_05820", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9804f74cb2db74e5e3cbba628bc3ecf03f9f9a8"} +{"id": "train_05821", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c0e96de117514013a45e883e5afcc4a549277fa"} +{"id": "train_05822", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2684f44cf4ea522415d5342aadeabbec17f5db6f"} +{"id": "train_05823", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b4adf452929e279ff72c3b93a490644f9e9c9cb"} +{"id": "train_05824", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "auditability", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c52ba51cf9742672c691660acb9a14afb3d8267e"} +{"id": "train_05825", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15ef2cd3f7cab371e65d8d610b80c0a94128b965"} +{"id": "train_05826", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c856b68378e0630e1c2953ea616c9e0732c383a3"} +{"id": "train_05827", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "auditability", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dde973614d1f4a1e1ec65e368070e64e332e08ce"} +{"id": "train_05828", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "942a7a28460223375d6da5bbce69bd29c8749c49"} +{"id": "train_05829", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cbf89de40302f6fbf9fbe99ae48b5217cb7c411"} +{"id": "train_05830", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "248adb0fb38147e3c02d32519e5d5b39add2ff97"} +{"id": "train_05831", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "tooling", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c399393562783866e5470a58e9092f0e1c350121"} +{"id": "train_05832", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "ci_integration", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eda04a25e5682f4f1a0a5e5e7f6f71544a708bdc"} +{"id": "train_05833", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "125fb6d922bc80b66c5d0f6f72953af75925def5"} +{"id": "train_05834", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38222fdefa4c18953142907bfaa9be3a863f4086"} +{"id": "train_05835", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f56bed9b7ce6e49092da83a6570d63af613bd8b8"} +{"id": "train_05836", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0bd124ee56aab7f64d0d503223b0b438c024025"} +{"id": "train_05837", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3309106fa438ea271db0bd0c2c11ab521a98b40"} +{"id": "train_05838", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0104e832460ae2456d6ef28d19552dcb78342b71"} +{"id": "train_05839", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9a367899bfcaa508f10fb9bbb0eee251891719c"} +{"id": "train_05840", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tooling", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8267abd0e65ba53656820a186f7b5bb14539215e"} +{"id": "train_05841", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "evaluation_metrics", "tooling", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a504ee8f25e7a00502418fa9e49a698bfb0a0a86"} +{"id": "train_05842", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "123628dfddaf664be3d686f116e8b926d6b8aaa1"} +{"id": "train_05843", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "security_gates", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "187b90fe7856efa2e5757a45818c17b674b52a1f"} +{"id": "train_05844", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f50af981a6a646f109214aab81bdbd4d37cf30aa"} +{"id": "train_05845", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "auditability", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "991f7b7ce2e0eef77abe3cf33c777ca48f9ed8a3"} +{"id": "train_05846", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27027e6619793c53b12f096f8bbe5ef71bb11e24"} +{"id": "train_05847", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "auditability", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d258ded644a58397413b4022b02763b8f40d6e3"} +{"id": "train_05848", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "699663e1359c37be749705cbe447c9d45809f05d"} +{"id": "train_05849", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be5383a2f1d51c9d89a05243ff2707154859e8c5"} +{"id": "train_05850", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a200c501fb823f68d1300f2ef1219f5fcfd4b9e"} +{"id": "train_05851", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e26c96ce4caf48375bd5f30c4866c7d8e0e0133a"} +{"id": "train_05852", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "documentation", "auditability", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec75424372fce1ccb52d97277b1ba1493f8a5946"} +{"id": "train_05853", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffda027be9afaf3bf6725024c8015d80994b7bf4"} +{"id": "train_05854", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "security_gates", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63ba6406e4f6c575f2eb3c3375ad639e527bad95"} +{"id": "train_05855", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "cost_latency_tradeoffs", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e14ecde58a422005bdc472d3545868337ba5e3d5"} +{"id": "train_05856", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "ci_integration", "governance", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a876b74dd01011ecbeb163ebc158203e3f92c72a"} +{"id": "train_05857", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "reproducibility", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7aec61485d0f8d8ff638b5d56f88ae6666a75c02"} +{"id": "train_05858", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c9c6dbee42d6b9909838977c805e69ff2e9ca9c"} +{"id": "train_05859", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a418b363c637ed3e16c7678f2dad36ed537d74a9"} +{"id": "train_05860", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tooling", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b8f4494cafd9bc7a2ce412bea2d2841c8ca3c5d"} +{"id": "train_05861", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "878c177b2628e217737d468f28f18273ae9c9d07"} +{"id": "train_05862", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ccdf88dd7d15145bf87c5b80de7e1e2915e974"} +{"id": "train_05863", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdc3032216d4a5ba022260ecd5075a3621586ec7"} +{"id": "train_05864", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37a9d0f22d7879b19d020d1c430f9e5b1c46dc0e"} +{"id": "train_05865", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfae99c409d0dfbe2dd6ee6801af690c13c60912"} +{"id": "train_05866", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "539d30f19f9f9f3c61ce6f9f3c0be1aa86c21b11"} +{"id": "train_05867", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f53be0b5288d2a747575a5b78eafa08dcd53a47"} +{"id": "train_05868", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5891b30153211125ea811bb0460c83c6b62935ee"} +{"id": "train_05869", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e53d1cd3879a5afcef57c0d8b4c5a4ef2fadff7"} +{"id": "train_05870", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40b58efda403c3d1396cc69d7af272ded35fe1d7"} +{"id": "train_05871", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3cc2a55268e8bcd9aced96c8854a7a83a4553b6"} +{"id": "train_05872", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79f4c0cd447a9e68031c9f2cb3ef4cf6fb13f897"} +{"id": "train_05873", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "repo_scale_reasoning", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d52724f03cba70cba94ba569e094b757490a595b"} +{"id": "train_05874", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b216c4250c72d049ff63a6f35c169868fdc8bfb"} +{"id": "train_05875", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71dabc92bba7dbf6a3c1d0fc8c32e43385aa5f50"} +{"id": "train_05876", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d5c5626db9b5971c70e5949244e9763ebf4fa2a"} +{"id": "train_05877", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "security_gates", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff2ede8739ca7cc0059cc65860d8f25d52791d9c"} +{"id": "train_05878", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "security_gates", "auditability", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d99606b0cb9adf70735f0f29f668326f324febf"} +{"id": "train_05879", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "documentation", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "edc6cfcc4cc8ac7a6fb43bf90a61fa0cdf04e289"} +{"id": "train_05880", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0970c907721c951b5ad61275587c557069bec019"} +{"id": "train_05881", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f429e4d8a10b45ca0a5912ff068eeb3907c1e907"} +{"id": "train_05882", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "ci_integration", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5714716767aab0ee517bd8b514a94a752f15f569"} +{"id": "train_05883", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0aaea0a23f4b05f3a204dc94e114845b39d46729"} +{"id": "train_05884", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0dfadbded1a4dc205a07b9324052831e89bc7405"} +{"id": "train_05885", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d4583d24f5aa5c7680d4110d387da9eabf02a5d"} +{"id": "train_05886", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "security_gates", "documentation", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bf33268f78ba19c54e7afe07c561af60e7f3628"} +{"id": "train_05887", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ebf110f3a4558dbdaef4517a0e8b0087182cc46"} +{"id": "train_05888", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8740203660241f17856d96222b44ab50f44dd244"} +{"id": "train_05889", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "964251bcb2ec9dd3d931752262d0cb0264abd797"} +{"id": "train_05890", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2531fff7fa9e74469635ae907a72e2151bb525d"} +{"id": "train_05891", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f2bc0f28750c16f0e6c92bbdfda334e84bbab8f"} +{"id": "train_05892", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "governance", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d39782139042eab69609be5158aa007297725857"} +{"id": "train_05893", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58f753d27a724d912a1a615b9c4e90b3ff697cf9"} +{"id": "train_05894", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b601357f95d21ad5ed49bd59d71627125e2f379f"} +{"id": "train_05895", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcb258de2e6cd99b05540b06426cdc6758b37962"} +{"id": "train_05896", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ee444e6e79708fac0690cb0a9e17300d37b5955"} +{"id": "train_05897", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "repo_scale_reasoning", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "886e4ed0594e7729598eb43c309deb90689088ea"} +{"id": "train_05898", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a1b381b889cef02cd7b096d4ad8d589d51011b6"} +{"id": "train_05899", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tests_are_truth", "tooling", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4bb26ca493bd48a313b02edc47c393fac6d0018"} +{"id": "train_05900", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "422c6ed5e6a1d5de88e878b199e09680922fbb0a"} +{"id": "train_05901", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8bac355c26d828661d3872576419df977448ed3"} +{"id": "train_05902", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b970e5694ac346f3af910b118a6b8a77aa4db3f0"} +{"id": "train_05903", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5cdd3ef8e55d163ff12350e83172b99c717edfe"} +{"id": "train_05904", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "documentation", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c708ff8f1990e86310fcc0619bbec3b2ec5ad0e"} +{"id": "train_05905", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac8b06856b6646d55b0727d8fe570fac48cf3fbe"} +{"id": "train_05906", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "ci_integration", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "026210b43ad7cd34b632692c3ccdc7af69215201"} +{"id": "train_05907", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "214313c47b66218f466ec151cd7714113228181f"} +{"id": "train_05908", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b532afd9cc0f6e9ccccd461048ff45b80882690c"} +{"id": "train_05909", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c464d2317e6f0785ff9ee15cec3ca2577ae8a498"} +{"id": "train_05910", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "governance", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9de37f70427580b12ea23b7907404f77c4c89465"} +{"id": "train_05911", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1faa206fc53c97cc165ded062a95ed3e5689f1f"} +{"id": "train_05912", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "documentation", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02f84503e746944627b680c53e59a15700f70018"} +{"id": "train_05913", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "tests_are_truth", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c6a28dd93440653576701257d8725077ccd6e48"} +{"id": "train_05914", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "governance", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b13dfec876a4f30840bde516386dc6f4dbdbdef"} +{"id": "train_05915", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcaaf63dc369605676aa781557306e1083e62a28"} +{"id": "train_05916", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "repo_scale_reasoning", "reproducibility", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bec1e9f26758465b4ad5074c8f3ca7f77f727bb7"} +{"id": "train_05917", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35efbc6a6742a058e4afde64678334d4d1144e33"} +{"id": "train_05918", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "documentation", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4234f9d08e8ea5ecc560a598d495346301ae881f"} +{"id": "train_05919", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tests_are_truth", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03545badabc72deddaaa02eca279212c8cb197bf"} +{"id": "train_05920", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f98d01b66734b06982287164f084553c3d29772c"} +{"id": "train_05921", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "ci_integration", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c86c240e75e63c19b1c818ad7586d098b8b78992"} +{"id": "train_05922", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "governance", "documentation", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84aa9b68e3b9f00f17645fbeb3f0a46594264ada"} +{"id": "train_05923", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "evaluation_metrics", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8efdc44d13e7e46de2e8c72943d0a63bf98d882b"} +{"id": "train_05924", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "213bb6bfc016a1436411f830fa772820bf1b4cd2"} +{"id": "train_05925", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a53d0a39b896b4da3a23ef0260b169ceaab0adaa"} +{"id": "train_05926", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b20c61e424cf541da605cab57295895e8cd60d73"} +{"id": "train_05927", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41410ee3ec4f217eac500691593e8f8a4eaf0eed"} +{"id": "train_05928", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "documentation", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7d2388377d0e1bfc5bb9c073c971768bbdb49bf"} +{"id": "train_05929", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "804d4cb051d59705cd9a802898e22e8f13407ed9"} +{"id": "train_05930", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "tooling", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6a70545ced29ad791d68bb3f3edd229a94489af"} +{"id": "train_05931", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "021aba1f41380da596fd2d3e50dec4240b3f6607"} +{"id": "train_05932", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ce247574c925702f8ddf7b324b0742d4ada6b0d"} +{"id": "train_05933", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "documentation", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6296af3eaf2e45058a13090085dbe3dcb71e0626"} +{"id": "train_05934", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "585670101266adb4dfa3360a0f001299935fc190"} +{"id": "train_05935", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7885b35c2ef490c05a35798ff3db75f0981e8c0"} +{"id": "train_05936", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4e5192e8d9ff011858a6b8940017f03f6860384"} +{"id": "train_05937", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "288f29b1bd60e76339565134f7bd1c98f03ed2b2"} +{"id": "train_05938", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89f61fd0aa7c2288cbdc1e521aa5d8ab17c05080"} +{"id": "train_05939", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "security_gates", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1158260df36c7c2462e7bf2f508cbcedc1f1f80"} +{"id": "train_05940", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcd851eb168fd2047dba9e3cd7c1905259b3e969"} +{"id": "train_05941", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tooling", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddff4f14c989a00c25bac6aaf2243bfb6ef74bcc"} +{"id": "train_05942", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42f1ad92527acf2056f5b2a6988c26cd6dcd119b"} +{"id": "train_05943", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44b0ddd5101a7c78b6749185e54792ea63c7dea6"} +{"id": "train_05944", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d36d17606321f4b601195ffe150df8001d382da0"} +{"id": "train_05945", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "documentation", "governance", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0347aed9169f6eeb25640e99dc03a8bb5801656"} +{"id": "train_05946", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "tooling", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b1d2cfd8ded336aa279b282cd9ad34d669bd896"} +{"id": "train_05947", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9635e598182370cc87840065a4a289773177e48"} +{"id": "train_05948", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "052310873c2d24fbb2893b74ce11296dbfff39f0"} +{"id": "train_05949", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "documentation", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49f8c41be431b68ab1b163316cfe4896ef844f00"} +{"id": "train_05950", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2d65370b816d103ae59596e64a63b0ad380d632"} +{"id": "train_05951", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "tests_are_truth", "governance", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23422ac6a94a4d86f98b54934196ba215ee346d3"} +{"id": "train_05952", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9466d6d9ecbc09dbd44b906545c6995007495a24"} +{"id": "train_05953", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tests_are_truth", "auditability", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3437ac42cf0f35e9c3c3898ae73211026352cd55"} +{"id": "train_05954", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "auditability", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b3e3a7e979b7a897aba87f8d459fcb0a103a36c"} +{"id": "train_05955", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fceae268ce611d1920122539cd25805ad8f2720"} +{"id": "train_05956", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "918d0744b72cdad74711c1391fef3aa2704f18e9"} +{"id": "train_05957", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tooling", "auditability", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d87e9024f4e7fa95e57469b7f4ae8f1c450e51f6"} +{"id": "train_05958", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "documentation", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7d3d86b59f01f5409afdba722fc48b93b4f504e"} +{"id": "train_05959", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdee0bfc761cbdc08513d0ece086225f1a50c734"} +{"id": "train_05960", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f05e991711eebf76e32fd873e81e1a7a236bc45"} +{"id": "train_05961", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "reproducibility", "security_gates", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1e4e45311bd08537dc5efc8906c4a41b6552c96"} +{"id": "train_05962", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "repo_scale_reasoning", "tooling", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b279a04791cbfef89823e4c18fee0af5d538159"} +{"id": "train_05963", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cef0db8a1c68ffe0c8ca39c25e5fbbde1eab40b"} +{"id": "train_05964", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90a0890a724c599d4590ea13d6aeae857375c551"} +{"id": "train_05965", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48a93f43173578ff3d388e2cf962d6f95ce738eb"} +{"id": "train_05966", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79ab0f21ac374e0864572ca217a700f8b4b3f6d0"} +{"id": "train_05967", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "tooling", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff283c98d3d8a8d3bd1665f6e0eacef5e8b265c5"} +{"id": "train_05968", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2d66175c49d98f2491e806b365fd77ff0662b24"} +{"id": "train_05969", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "ci_integration", "auditability", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1bda2f7674e79f258cebc79df09fa4a7fc36314"} +{"id": "train_05970", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a860e9af1c98575d8ca84c20be1f0a3cc4c05a4"} +{"id": "train_05971", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "780456c6eb0b18f50df8e5e0aaaecc2601550dd4"} +{"id": "train_05972", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2ac29bf14d8921a176f2e264947045abf4b1ccd"} +{"id": "train_05973", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "tooling", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75b1b9d4ce67d3e2ac4d5d19c0fd5f4a8e8545ba"} +{"id": "train_05974", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "documentation", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58c9bb938d7329dcbf75ded40bfb1b55c408b1ed"} +{"id": "train_05975", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "679044c575fbcb6e4b462cb5919d4230cd61153a"} +{"id": "train_05976", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0139ffda130a334a816abe969fb2e307b53d0888"} +{"id": "train_05977", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "587266e7a2f86728d3d6937349de4df6373781ba"} +{"id": "train_05978", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "tooling", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f032e0ef0e2fd0dc053d8bf740342137b7799dd"} +{"id": "train_05979", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17dbea969d3744c642c61138421d7124d29fe950"} +{"id": "train_05980", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9199706d032336d99a83682ce320351629a7eeb"} +{"id": "train_05981", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62936cc5ce4ad14fd71627f11c8a4ee736f6d830"} +{"id": "train_05982", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26d8627909420055e975fd8dc04ec0200cb2f85f"} +{"id": "train_05983", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bc62326acc5365376bda1670d59403eceaa0730"} +{"id": "train_05984", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4ae3f986859324ee985953126efc351f7b76aa6"} +{"id": "train_05985", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd7a02e141931f340b6918f88765ce07e99382a0"} +{"id": "train_05986", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "527fc3df6954f080a1256fd19a8f3cc63b6482d5"} +{"id": "train_05987", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bcf21fd6ab52c21cc2adcfe2aa4ce47bb7a6a38"} +{"id": "train_05988", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "documentation", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc02682ccd83cfbf97f47121bdde6258bdad6ced"} +{"id": "train_05989", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54aefa45c862919c1da4c239bf0d8f24d5542570"} +{"id": "train_05990", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62e32ee3d67e4e3ce5a37ad683d23a0085d82d5c"} +{"id": "train_05991", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c67501a550541a3cc7737d8bbbb33a1f694c5ce"} +{"id": "train_05992", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2987fcfd08395894c82948be8f8310e60a87cbb3"} +{"id": "train_05993", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "security_gates", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8cf61144cb04f34792e79c63110b7e6a773285d"} +{"id": "train_05994", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9f8dbfa52a94a333c8050a16c705c3fe6460ccd"} +{"id": "train_05995", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "tooling", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dea06121061d2becaf37f82af694a1873c763e96"} +{"id": "train_05996", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eed67ba6cb3af11e05409f18b290cc43f95c74be"} +{"id": "train_05997", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "ci_integration", "documentation", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "587266e7a2f86728d3d6937349de4df6373781ba"} +{"id": "train_05998", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "security_gates", "documentation", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ade4f69f450e73c4d9a0f6b660e284ea5b113df"} +{"id": "train_05999", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "663414870eca80f8f6b6c851993f4543fa066dc4"} +{"id": "train_06000", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "documentation", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a394575e27f3a40c4ad08bc997560feb6d92ffd"} +{"id": "train_06001", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9bdb3d5260b3e60dea0619ef08bca3352cb2d6ae"} +{"id": "train_06002", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83f90eea36104cf5d2a155d9875ebc13ac8c6a95"} +{"id": "train_06003", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ff94c75f34d6585bfc3fee6cd18bd881e7b894b"} +{"id": "train_06004", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "546e1744ad7444806b8225d5d4ed4e925fb0f080"} +{"id": "train_06005", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "documentation", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e50c942bc705d01435dbe78b20fa917f8971768c"} +{"id": "train_06006", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "reproducibility", "documentation", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10b4c9c5daaafc5c1ea07dc01958ebf54a90429a"} +{"id": "train_06007", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91b8f0d135ea3986fbd91ab44ffe1e26ef82a0b8"} +{"id": "train_06008", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9bc102617e5f8a03bf0787a6c7bccde04f36951"} +{"id": "train_06009", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "auditability", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52a634069a9799c4d7d2b247f6385543fa6c5e68"} +{"id": "train_06010", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a4a5048ca10e7709d6df5853b33ca894d2130cc"} +{"id": "train_06011", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "auditability", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f61e2a2a02de9a92333c23196ad0ea7499150c7c"} +{"id": "train_06012", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "reproducibility", "auditability", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8fb131c42fff326761891678cc632aa24a7aa5d7"} +{"id": "train_06013", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3d90aa766ba6a7e52bee82b6a72039f3f791844"} +{"id": "train_06014", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "728b4c1da6058cb2b2e77f7c350ac755293ff00e"} +{"id": "train_06015", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92a77c14d72843a238787ba51c64eb059684f951"} +{"id": "train_06016", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cbab1122d6b4ce7af40050b796a5cd54380dbb3"} +{"id": "train_06017", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tests_are_truth", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc62b2e8b6593948ea77e4836298fac98176b4d1"} +{"id": "train_06018", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4268d9700a9b136d5b85eec8d8f8d71f25673ccc"} +{"id": "train_06019", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "governance", "tooling", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4ac61ddde2e1488d65decc675ff5505a2b05dbb"} +{"id": "train_06020", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7942400ae274d16be096942845da94d5d9ad4503"} +{"id": "train_06021", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "tests_are_truth", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0da6a17b22abb23aa62976cfd09d734db4ca2b0"} +{"id": "train_06022", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "998f941d3ac23a65f9786237fc5f03f926445542"} +{"id": "train_06023", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8769e08fd84a6a38a8b68153e2849cfa26d5530"} +{"id": "train_06024", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "tooling", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "edcd5fe162b95ff34b5d2154b9e53581491c6817"} +{"id": "train_06025", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "security_gates", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c4a2580477b0deba99031eca3a243c8a2b3900e"} +{"id": "train_06026", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tooling", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54bcd8e3b42da83f28bf0bb492be083698f9320f"} +{"id": "train_06027", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1159d8b4d95ae6f24102acb993bc64363db00816"} +{"id": "train_06028", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2d8b8da0fb779b98befc862bd0e1093438bedbf"} +{"id": "train_06029", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "security_gates", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fefdee8be42e79e8278d741c7fbdddbeca012af8"} +{"id": "train_06030", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "reproducibility", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc9467f04aeeb5e23f6e8d9e26d1083e9a1511d5"} +{"id": "train_06031", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "documentation", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3ddec5dc810a880e21fa6b276a3a0af1b233cb8"} +{"id": "train_06032", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "governance", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55b39708e3f7996a7f7a4b299f8d935dcd5129b4"} +{"id": "train_06033", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "auditability", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29375ad93912c9f22bff34b837cddbaf79de6670"} +{"id": "train_06034", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03a2be210c9b91f55a3e156fc1daf32c0cb47331"} +{"id": "train_06035", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "cost_latency_tradeoffs", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8bb467fe16f0393caf8096ce3c4346b105140ab"} +{"id": "train_06036", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f82b4a683ebbf15b09cf8d20104c4b7d55c9646"} +{"id": "train_06037", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15b5ae78aff40c76c3f3982906281409dd0d31f4"} +{"id": "train_06038", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a571e55e066eb53ba2f6ea3bdb2ece7b2771388"} +{"id": "train_06039", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e519c5f9aa05b5dd8ac993d2505d86473b1624a9"} +{"id": "train_06040", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54de4d6d318669347357ff35cc9bb08e6b74df20"} +{"id": "train_06041", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d65e0764e75db68fb44bdf6138e216466d320213"} +{"id": "train_06042", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "acd0d1647511dddc208479fc91eb3e61569d3c88"} +{"id": "train_06043", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77be0753357ff2dbff9081fb7fcce13f9363f7a3"} +{"id": "train_06044", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "auditability", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc2199d999992d608f9f083b4a999122b72ff481"} +{"id": "train_06045", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af0a1ec68f5fb69c31aa0dd02f01824b6a810a07"} +{"id": "train_06046", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "reproducibility", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59dcca3e66246ac8ac42c5b92b4aec5632ff9f2a"} +{"id": "train_06047", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac1f9abdf028f35042561f4c553d8c657d480bd3"} +{"id": "train_06048", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "auditability", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0aa2072913acb43b1f1888ee5eecefb54851414f"} +{"id": "train_06049", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tooling", "governance", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d4605192527c9da0b6725f0defdb9835afad935"} +{"id": "train_06050", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c76d744c240193bb05b19877e0c1cfbc2fc12b3"} +{"id": "train_06051", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "auditability", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e4adde41f0f0b0bf139345dd64f0f0d1db956c1"} +{"id": "train_06052", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d49a371ceebd0303de13f82380e600c230ce84bd"} +{"id": "train_06053", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d037878618ccca634011752ef0e33d75219ed1a"} +{"id": "train_06054", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24ec06c71883dc99b78ee58f8e72880d3908963e"} +{"id": "train_06055", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecf43aab777ca4347221fae797bd13d58fac9dac"} +{"id": "train_06056", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "tooling", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "739a5277096c1b19c463a2bf31e4c2ab40ad037a"} +{"id": "train_06057", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90b12c967788efce0835ef7acafe077564c1811e"} +{"id": "train_06058", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "governance", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec00dfd4fc667ead8bd61bfed10ef77f5dc0e37a"} +{"id": "train_06059", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15006074fb4a3803a33c450c7deec2626e4e7cea"} +{"id": "train_06060", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a752d388e5df568f996be024d941bce436f9bbf"} +{"id": "train_06061", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b9329351117c01f446d547a33f47b34d52faea4"} +{"id": "train_06062", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53b60aa39a6dd4efa24a8cfd945bf015e2f364e7"} +{"id": "train_06063", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "security_gates", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4516c939e4c77ff5e438d83fc78e4b9ca02b93e"} +{"id": "train_06064", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49e2507e561ec6b9408a27bc260876c048ba8e15"} +{"id": "train_06065", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc703e5dc2a9783839261deec376dafa3ca569c0"} +{"id": "train_06066", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "001df8d625420b32aaba69ce9b2248774ebe543d"} +{"id": "train_06067", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "102b5b7a247c6185580f3764257368202ab17ad4"} +{"id": "train_06068", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "repo_scale_reasoning", "tooling", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a553d83a955a8a69c4b918acf2815e3085275d10"} +{"id": "train_06069", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f169b906d4293d1aa4fe36a931abf71780eb965d"} +{"id": "train_06070", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "governance", "security_gates", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa2927fb317c67ffa7f0ae93ba17d654751202c0"} +{"id": "train_06071", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "995bcbae683d43387bdadaa99a2bfe234792d021"} +{"id": "train_06072", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "security_gates", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b98c4e644f1de83a2c1c8104fae19003cb0a94c2"} +{"id": "train_06073", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa3469ae3d9aa43e59acc406a04d49f529f18df1"} +{"id": "train_06074", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a48b88a0b02911058e605dc453945aa3885395de"} +{"id": "train_06075", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ba27224adda7f3dcf70b6be91be72ca950a1b41"} +{"id": "train_06076", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfbdfb8f4c934b34dc7d705e240759a2c64c49a7"} +{"id": "train_06077", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8706c3596140063441bc438782a867bbcc66048a"} +{"id": "train_06078", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "security_gates", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "873c174b40a5cfe5bfc0c930c8dde9a1c6dbb405"} +{"id": "train_06079", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tooling", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5c8b637cb8b3ac5633542199eb0043edc6e9d6c"} +{"id": "train_06080", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c028a36719ffd2745154a9739303df4fba9854b8"} +{"id": "train_06081", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdcf1d35ac33c41dbeb28f6b0ba0ee350f665bc0"} +{"id": "train_06082", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad1b837a0063e946f214fe51aeed78fd81b02967"} +{"id": "train_06083", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "tooling", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1df23029ac49697b135e85d73e9ba31defd8af0b"} +{"id": "train_06084", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76f643fa289ca25a2a4fefe029af368ba1f598ab"} +{"id": "train_06085", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6dbb4c3e6e23a7fb9b7d5f3217baf35f38a59449"} +{"id": "train_06086", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "ci_integration", "documentation", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0e86444332bf0d5858d609ab5b5ad0f3cf355f9"} +{"id": "train_06087", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3df3afa142d49d861a56572212e6fed3098d16f"} +{"id": "train_06088", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a5c405c8085942d71249383840899beec1aa417"} +{"id": "train_06089", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7046562a89f3ccf949aae9d74035aef0f435870b"} +{"id": "train_06090", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49ac79c34a8cd0ca0abfe54a1146fe411d296de6"} +{"id": "train_06091", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0dac6ca89a26bb716bdef4bfaf2efdee80498ce"} +{"id": "train_06092", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "tooling", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4615d90119f5fc94507145e43cdbf3f3a1cde40"} +{"id": "train_06093", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77d4ff7be59b38e1d883411227fc1f19cd574ed1"} +{"id": "train_06094", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf2c78e0ba690df7ca6fb58c1eed7d72599fad79"} +{"id": "train_06095", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62d2617056c2b717d1866d6fbfd5a769210fda89"} +{"id": "train_06096", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d2af4ee04bc6b9b417c4c580b103643b40ea49e"} +{"id": "train_06097", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc11e6fc8110586b0f0100862af05b6ce8971adb"} +{"id": "train_06098", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "security_gates", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfdc02ca850dad0c269e6c96386fef74c3c02404"} +{"id": "train_06099", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8fb5bd816e1dc4f331d2afc070093541e6274ec7"} +{"id": "train_06100", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5834119c8677d0f933dc5386bc31bf84885aba8"} +{"id": "train_06101", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1b71de7cc98980b6bf8adcffcbe2fcc8367ba53"} +{"id": "train_06102", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a1c5972068f78bb736596daf1569b0ccd11f38a"} +{"id": "train_06103", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f7b2f8730c8ce252fc697ff5f3aea7e498a08ea"} +{"id": "train_06104", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2cff725c5c522b7abb6f9b42c044083c94faf55"} +{"id": "train_06105", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd4cd14560e9e157af7ae18992e7fc78c99add4e"} +{"id": "train_06106", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "499415d3ad8f1f9fba2f6899a83a6277af77cd2b"} +{"id": "train_06107", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "security_gates", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5035575bb8fac13d795235622fda8ae6c003a4ea"} +{"id": "train_06108", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "277ce459236f361b5beb84b1c9e706ab6155580f"} +{"id": "train_06109", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tooling", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59531f312b0c45e45a1de03bb5da62513f495815"} +{"id": "train_06110", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "685e7067faeaa2bbbb7bfe39ecebabc184372af5"} +{"id": "train_06111", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8538f903e427d1dee014983054cfaf2ee440c031"} +{"id": "train_06112", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f1950202571672404226cb4b67328328074e271"} +{"id": "train_06113", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4432df4713420fbe525b1f0d5a89d7f2837bf63f"} +{"id": "train_06114", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fae6758a4cea977c8d3dc2de4146ab18b0a0b444"} +{"id": "train_06115", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "323294c469cc9ca6547d37459fd7133f54fe4ad1"} +{"id": "train_06116", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tooling", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20ea7fff40772c6a92415b4f9d1adb8b11afac27"} +{"id": "train_06117", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1681229f4834c232534e3fc5bd2bdee10b88a88"} +{"id": "train_06118", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cb3fd1a71ec7bc2a28d8a5c5fd527edae44cefd"} +{"id": "train_06119", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca5e5447cbbdf790849d38141c4ff59f07e5756c"} +{"id": "train_06120", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f257ad17bc41412e98a77de34777c7edc31b7dd"} +{"id": "train_06121", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "governance", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "883d70c424f71b5e7839bbe22eb2b372de701caa"} +{"id": "train_06122", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tests_are_truth", "security_gates", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b419d10651029c143eda254019ce44c20c97ff0f"} +{"id": "train_06123", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c88e18bd7a18fc654260d0dc770bef0cfc76522"} +{"id": "train_06124", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "ci_integration", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12673612d6410eb70e6cb706b830bf14e0613258"} +{"id": "train_06125", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4319687a01d7d0888f39381fe904c365e99dd822"} +{"id": "train_06126", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0be724a038ae6027d267dfffad7dedd1b818b73"} +{"id": "train_06127", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "ci_integration", "tooling", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "470bc483b88180241146e8ab4c6779094372480f"} +{"id": "train_06128", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "reproducibility", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "284ab27e7f3883f53170f1c6d3d4753a46afca58"} +{"id": "train_06129", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "reproducibility", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ee444e6e79708fac0690cb0a9e17300d37b5955"} +{"id": "train_06130", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce8ed1a8e1744af4c552b752fa2332f1b654ee9a"} +{"id": "train_06131", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "auditability", "tooling", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "554087d6b7bb2dc593d692054fe4cf3be1479dac"} +{"id": "train_06132", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d131bb3522c1ff551a7bf3ec74d60337b8092183"} +{"id": "train_06133", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "governance", "security_gates", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c6e4dc7059f84c3105265d5fe7b8091cc6213a4"} +{"id": "train_06134", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tooling", "reproducibility", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aadd17fdc36d49c4f794117dae6f3019b39df5fe"} +{"id": "train_06135", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42fe14bb145bbde31a33af4b3cbbed84c26a05db"} +{"id": "train_06136", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b65a7215af0f4a85324fcc50c1a025f7ac63db61"} +{"id": "train_06137", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62bdb40fb861a019bee08c2564234fa5e3dfdd0a"} +{"id": "train_06138", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2374b84bec676f7428b29c648981788787251ff6"} +{"id": "train_06139", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf7b38408227212f9a4bf566588e721165cf7343"} +{"id": "train_06140", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c8e9b6f07265c754e6050c502c55e7249c81481"} +{"id": "train_06141", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "evaluation_metrics", "tooling", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "428cc249c77e576fe63123820478f395d3a7bb83"} +{"id": "train_06142", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec5bf6e59dda649e20fbe00910d19c9b1850affd"} +{"id": "train_06143", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80c9cd241f1e3e86653e5dd2f1599ecc7a47b1b2"} +{"id": "train_06144", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf52209e60c3052f139b5b2f9a0fec25006a4415"} +{"id": "train_06145", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5e1ddd4f579f827d243fce64a6f17064229b85f"} +{"id": "train_06146", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85036cb3a7b8b85f35e095f9a309c502e8a003f0"} +{"id": "train_06147", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf31a00501e73e7f74c47028c42c4e06689b540b"} +{"id": "train_06148", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "evaluation_metrics", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b95ef409761007533aa41308e4444e23f73d7fa8"} +{"id": "train_06149", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "governance", "tooling", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdb1fb1a691d698bc256fb2d75995e55ef3e9ea8"} +{"id": "train_06150", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da45109f17322c70695b6f75e3432bc5cf18b2c8"} +{"id": "train_06151", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "repo_scale_reasoning", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b32c04cac2b50e850a21cb1913aa8a8fa18450f9"} +{"id": "train_06152", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "documentation", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "299b4b81948b2db0d72b90f98c6d022a0e01240c"} +{"id": "train_06153", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "documentation", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "672e372d4c8bd960801cde2eaa9959ce23c0a652"} +{"id": "train_06154", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "597aef5cc136db30b3712ac5edb4705cea49ae78"} +{"id": "train_06155", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "654e23ec3772c9f2459273b4adc9ca70ebc0220f"} +{"id": "train_06156", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a77a925bb6c1e76734f285a0d975189d6548088d"} +{"id": "train_06157", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "691dd951107cb7fc9e617ab60d67df3466e54451"} +{"id": "train_06158", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "auditability", "security_gates", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9de8ecc0d1bf2788663a05cd3245d6feddfd20bc"} +{"id": "train_06159", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebd9f83e8cdb35cf0220eb9c4d2663d43a3894aa"} +{"id": "train_06160", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af5a3f367ab8f32632d3ef367551a6bbfa8625b2"} +{"id": "train_06161", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "documentation", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a269d0110434df8cf205bc58ad780d611b345f3"} +{"id": "train_06162", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "633429ef2dd363c1b1e1887b77a52b8a25611493"} +{"id": "train_06163", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "038c75543f531aef0538b4af74ff4e4a1d94af0f"} +{"id": "train_06164", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2277b9f0cd23c4d35459f7b70b0c2fa0c04e6832"} +{"id": "train_06165", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "365f16a31a500ba8c55b50f16019dd6e10ad3808"} +{"id": "train_06166", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f57d4a3c37c622f45436d50c36ffc317a70ee8e1"} +{"id": "train_06167", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "security_gates", "reproducibility", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb8fbe885dbc69761128c9bb5d088ba5a5d67647"} +{"id": "train_06168", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02b7ff1f4bc883f22f9599902028793e823397a3"} +{"id": "train_06169", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "reproducibility", "documentation", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca02dda599706606b042a79824fd1ca84ba97b11"} +{"id": "train_06170", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbb464ee073a227d221f9a6c61b1ccff683b3343"} +{"id": "train_06171", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "ci_integration", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0ad47f24e8dbfc6a8c10726d7ffa97887202f61"} +{"id": "train_06172", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "reproducibility", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86725fd5d0b7c729f2c8ea630f0e65bc2b396d96"} +{"id": "train_06173", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "347e8472bf06e7a8afcfdedc18ee4492e1ed9961"} +{"id": "train_06174", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ae01491838a155f91980c42a857af540d9e840b"} +{"id": "train_06175", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11357d13fe65b51ab0e6670bb3a05ebf9c741f4d"} +{"id": "train_06176", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65f5120ce441973df1e1abda48156fdd0ffae285"} +{"id": "train_06177", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "reproducibility", "governance", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd092b753e1edf209bf24f0828d317ff522d7cd5"} +{"id": "train_06178", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5148af9f0bfe32d906baedb2d8693d38910fd4c0"} +{"id": "train_06179", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "governance", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9ab7c4eb0392e9be9dacb8463df9f6563dd33b8"} +{"id": "train_06180", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85a5ee0191f5b4c83cc903b3209d4bcf3f9742a5"} +{"id": "train_06181", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "065d335d6e2646c5816962daf5e8b50475b62144"} +{"id": "train_06182", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ea35580e409eb99bb5a99e17254c1ea1ed9eb32"} +{"id": "train_06183", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58f4419d58b3045e5593a2578a96a6e67f1e1078"} +{"id": "train_06184", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69a4e68b1202f1950042bad8a596db7a8219896d"} +{"id": "train_06185", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8632cf9c0c95ad8b36eedf01d5e226d2ddaed1ec"} +{"id": "train_06186", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b9495478c6ec53b80cb5ce8e6d92c24e97fda8f"} +{"id": "train_06187", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a45dc7a6dab05cb3057257e781cb449e84cfb55"} +{"id": "train_06188", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "auditability", "documentation", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d7c01474ab1910eab9082d7233859f9f3b19de8"} +{"id": "train_06189", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "governance", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be293c127632777fd7726529c98d6d108a453499"} +{"id": "train_06190", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03661cb3a1d0d03e48024a6f981f0f161fd026dd"} +{"id": "train_06191", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b1abed2b79995f77a6ca18f6853612bfd63df32"} +{"id": "train_06192", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17270cac3b044e72cc407574133b036cea2b66a5"} +{"id": "train_06193", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7ed5cf9c0c084f399d7eafc9749d5de38ce321c"} +{"id": "train_06194", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56384012a60417dbf7d1e8191b3acdbecd6d6151"} +{"id": "train_06195", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5998b0a376ee36093ed8552bddd786e47302e07f"} +{"id": "train_06196", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "auditability", "documentation", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1dbdee4368083d91cf2f23a6642379d5cf3f6133"} +{"id": "train_06197", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbac4943760a5370dffd34418bf8d347ab427617"} +{"id": "train_06198", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0d383da96da6fbbcc623468e3254ebc8f1e83b4"} +{"id": "train_06199", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe8e7eacaa608d72cedf07f7fb2f9db230a25505"} +{"id": "train_06200", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "ci_integration", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "325de52976c5b2345d9456b94259b97044cdaf72"} +{"id": "train_06201", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tooling", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a0a0215a163b0be574858bcdfd0dcbb0ac13459"} +{"id": "train_06202", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "security_gates", "reproducibility", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cbdb8004db78d99fc3d359d45c546b3b0cc5cc1"} +{"id": "train_06203", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "documentation", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae07a5ebc590693fc860b82f569ed4a648022d75"} +{"id": "train_06204", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "tests_are_truth", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "355cc3fe6f077e6ab21eb44279144bd47b631f03"} +{"id": "train_06205", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "security_gates", "governance", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92a37f59bbaa9cffee440d96c6712ce44a2768e0"} +{"id": "train_06206", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fd705d82a35bf58f21c591b03c88a0734c13576"} +{"id": "train_06207", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8660afb82de5d21ddb32bcfafb31f32ccfc53111"} +{"id": "train_06208", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d1888160012f08d137d146e62ea1fc89ff12abd"} +{"id": "train_06209", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c494727b3ed7c1510041317ca779aeb800f4c5d2"} +{"id": "train_06210", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "documentation", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73d53c9590ee4deeac43dc281a4d088c8effcf82"} +{"id": "train_06211", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbd69dff66da5c4ca3bb87beea7b45a854349276"} +{"id": "train_06212", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ada05e749580b6e74fd2e1f70e2d6bbde5dbfc45"} +{"id": "train_06213", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "documentation", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d4d214bd96dc715b26e1f4c7c98cbc69e8cd667"} +{"id": "train_06214", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "governance", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89dc102b5de1777863106856bf3753d8502d27e3"} +{"id": "train_06215", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f99659e763e465e95cd4ebed0b1078bd815e62ac"} +{"id": "train_06216", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11d3e256948e6a888d14f8fc823a84ac9b073271"} +{"id": "train_06217", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "012d0a65f129eb520064a2cecf68965f16ef8f2e"} +{"id": "train_06218", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f87f21434e2dba7c1bd097e4e7526466ca78804a"} +{"id": "train_06219", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5993ad71f417e23d6dd64453be4236defd0de693"} +{"id": "train_06220", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "cost_latency_tradeoffs", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ebf46b1d260d131f4edab8436b04855abb14c85"} +{"id": "train_06221", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c77d139717af6e8bf76035afafacc550e0d7967"} +{"id": "train_06222", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "888344fc78028f4d6d979f6b6ae16ef690ab8c84"} +{"id": "train_06223", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "documentation", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7f536c8d99c8891b546a248bfe540e96af8a8e4"} +{"id": "train_06224", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0760ff22af4c5981abcfb93c05c3624979c45a99"} +{"id": "train_06225", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0bff789c9b3abbb061fb19c7efbd8f8bc7599d6d"} +{"id": "train_06226", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06e51241e9deaea4dbea2ccb3e53e0f9ddc3536f"} +{"id": "train_06227", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6027949719b1ac6293ed2a1657a0fe5a9c5aba96"} +{"id": "train_06228", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cffca0f7e75c59824856f4aa2725a5df41ad8ba"} +{"id": "train_06229", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4da811513241049583322cc9cfe4e8016cc75b5"} +{"id": "train_06230", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fe53fc398818e88c72a86ca11180f438aef4e22"} +{"id": "train_06231", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59e23404751c9149fcc8af5cf103ee695e8ab37b"} +{"id": "train_06232", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b4fc84d4e5aa5e8a82d339d66708ab3c8f6adbd"} +{"id": "train_06233", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "276809749c4d6f5823d2fc164fbae6c0137585e4"} +{"id": "train_06234", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "repo_scale_reasoning", "reproducibility", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a53150fc9fea421c3bef73e76f9adf88d5944ca1"} +{"id": "train_06235", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7f431d7ad246451e014e779fcaa963220d7c032"} +{"id": "train_06236", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e34a100e64c6b2f8725e76edb255f082d211e2f6"} +{"id": "train_06237", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "documentation", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3995765778210767e00f68632e4309d4b98141f0"} +{"id": "train_06238", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "254b093138b9a947809e21767e40165814e8a13c"} +{"id": "train_06239", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "tooling", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c06a54422052f130431c8428bc3c381085a6d7f"} +{"id": "train_06240", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da15d1bfd9c885e2b09e2e99b8b7e21b01e0c011"} +{"id": "train_06241", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "governance", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1551e3ea505649730115f3b1b5aeb0ba3d57a1df"} +{"id": "train_06242", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c06be94d11743f2eef11ab52b763607bcba2c2c2"} +{"id": "train_06243", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "governance", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a78eb017504a5016b19d707d5cc9c474f9b08af"} +{"id": "train_06244", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "documentation", "auditability", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a0b3c11d47ee9af33513db3725f97ac11e31098"} +{"id": "train_06245", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42ff26ae17345ce63726ac4469edefcc9b9af93d"} +{"id": "train_06246", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "783bdd3bee3c3755eb85787937703227be0f8709"} +{"id": "train_06247", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cc9dd22a185119abb3d8ac3cd81f3cb481b2abb"} +{"id": "train_06248", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "tooling", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de29a1c3bc97391a7178cfd71afd255493e8f65c"} +{"id": "train_06249", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "governance", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7e79ba9d9c225f76b8cb6b6a2fa0cd7cb748f0f"} +{"id": "train_06250", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tests_are_truth", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a822c275e3e8807f24808819aaff8c76856d43aa"} +{"id": "train_06251", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "security_gates", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f343ed7c97e36619a824fc266a521187ce887cd"} +{"id": "train_06252", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b98c4e644f1de83a2c1c8104fae19003cb0a94c2"} +{"id": "train_06253", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d9fa99b5355d6bf1bb9694f2e824ac898d82f72"} +{"id": "train_06254", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "governance", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1abd9b30a5e70d95a928223501e3c18019315d6"} +{"id": "train_06255", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74a929bbcc53b8ffba75e05c074ec4d56f65bc10"} +{"id": "train_06256", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "security_gates", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b1676a0a99ced1d138c650f52d071298d88e9e2"} +{"id": "train_06257", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "evaluation_metrics", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3316b30b10750cdff4d1412a061833baac29f78a"} +{"id": "train_06258", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81fa390faeb47f215ce3820566122ffbeff20d51"} +{"id": "train_06259", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a934f052b00f26ea4139e5fa72bfcb96c7e09e32"} +{"id": "train_06260", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5ecebac2cab0654cf9eb6f492d4949f654ec599"} +{"id": "train_06261", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "documentation", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "febec1bf6802850bf844d69965fa7c9331b4fe6a"} +{"id": "train_06262", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01d06444a2f2adfba849c15664418c9216c528ca"} +{"id": "train_06263", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "527d53a418145c76b8dab19d9776fd49f8cfde1c"} +{"id": "train_06264", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "reproducibility", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb769ad27bbf7b27b918f24072815ae69924ca63"} +{"id": "train_06265", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "auditability", "governance", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2564c8fd01ea189ce128c912036ced60cc1e492e"} +{"id": "train_06266", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52c48b33267de16d2f2648633d8bc3d5ab9dec35"} +{"id": "train_06267", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "908e8b638a6ac874d6852b13a24cb8cdeeaaff7c"} +{"id": "train_06268", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tooling", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c087b4a067b451f249761440e9ebd7185c7354d7"} +{"id": "train_06269", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c664aa6c34aecce523195e770907f4f918333981"} +{"id": "train_06270", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2750524268a79692ffe21dd8880276942ac8d15f"} +{"id": "train_06271", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b279a04791cbfef89823e4c18fee0af5d538159"} +{"id": "train_06272", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "tooling", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "044f02f93d16e8c7871f59e0dcf17a80bd76b418"} +{"id": "train_06273", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91b3611879f8c123faaa01d8dddd77cf042336d2"} +{"id": "train_06274", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9e4e0faf7d4ec8c821f2febfe3d100e16af3b3b"} +{"id": "train_06275", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c1fb858fcd83621123ecaf5098a908901eda622"} +{"id": "train_06276", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8f0510fd4fc3531e1a1bb757c7b81ca251089b0"} +{"id": "train_06277", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cc1636e48ede569b39ae63a3ae547def791afbd"} +{"id": "train_06278", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1de13f59dd31c89e3e62055eb3357c225637d4c0"} +{"id": "train_06279", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7905d90d3d9a3b4015730555e700956044c71302"} +{"id": "train_06280", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "security_gates", "auditability", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c12cdfb2f91f81e9881ebab902e7f5ebb37835d"} +{"id": "train_06281", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92499f2746e40fd916cfd355376c9a09c0e55216"} +{"id": "train_06282", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91722486214f4e38efe08c988f1757bae46711dc"} +{"id": "train_06283", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "369f8476612a04791f506b5300e7041e2f749506"} +{"id": "train_06284", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c63cba48739fbac651ba47f59dad4536680b3f55"} +{"id": "train_06285", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bdf5f454e32fb6f10b4b58e437135d728821527"} +{"id": "train_06286", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c64611c81f77f9277ef7a93fafbff2fd86c23ce"} +{"id": "train_06287", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2014c6be3dfda01a10bf5a2c5fcbad94803d700"} +{"id": "train_06288", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae3aecb9aab3a4be501ca0198cc1824f2459bc11"} +{"id": "train_06289", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88b785d3ea42f4f2098cad00201321b5092f43e7"} +{"id": "train_06290", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee11923a4737b2c062dd2c1ef26a3a8864e2ee0b"} +{"id": "train_06291", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "ci_integration", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66700c4cc3536ce26846c3b6f6da256d7c3096d6"} +{"id": "train_06292", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "auditability", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e93b10121ffc3b515a78307aa6d8338db0bf5503"} +{"id": "train_06293", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26cf12347d106443f2ef2b0d229d6dbe5350e4ba"} +{"id": "train_06294", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "tooling", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c514161a197ce05a31853cc883704cfd7d394509"} +{"id": "train_06295", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "security_gates", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd775a1d6964814cb1579740e21012492a465a1b"} +{"id": "train_06296", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "reproducibility", "auditability", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b264c661a517c1635a375d608ef0b61e8e56ef4f"} +{"id": "train_06297", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "548957317b48e1e6b8beaecdc438f4e7f154f63a"} +{"id": "train_06298", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31bcd8cb4a5b67258b98b1b49311ad20dd402c7b"} +{"id": "train_06299", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca50b7352d18fc5cb6eb519a3b0601249894b5cb"} +{"id": "train_06300", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e459bc3e15dd0a0039c251349a6cf52cef5b9a01"} +{"id": "train_06301", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c47ecb6f13ac3844e3737e556ab0ed66d042fb8"} +{"id": "train_06302", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f01ac8482e1cd19b4925b6ad40814b3b0a50d585"} +{"id": "train_06303", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "reproducibility", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "607e76087610319a619ee82d30ce9d313ec19ed1"} +{"id": "train_06304", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "tooling", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0cb104e5b61b33c373d5f9cf4e713406969f108"} +{"id": "train_06305", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3391a3c16325ca0c7344bc2d22662ab062f9906"} +{"id": "train_06306", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be2d4397aecbf5d27859c1379174952851d3f508"} +{"id": "train_06307", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "documentation", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "102217c9f7315a6d77157d529eda1f51bf82272e"} +{"id": "train_06308", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "governance", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1436857db7dcbf4c855f60dce6f247f1bf9f1f0c"} +{"id": "train_06309", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a47dfb1109831e4c40d3aa436ef72fdc60b48026"} +{"id": "train_06310", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4ef79b5b420626c37f8a020915993a9d3af648c"} +{"id": "train_06311", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca8c968dbe7e5d50af8145879e2a3d7c856cd161"} +{"id": "train_06312", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba0447e2dc9dd3b56e4e1c3fb8a5ce499b1ebc36"} +{"id": "train_06313", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "reproducibility", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53a111e7690be14aef1497531ec7291bd8aaff1e"} +{"id": "train_06314", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "ci_integration", "tooling", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c9e1515bd2c1e2596491460aca7083b66d20b18"} +{"id": "train_06315", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tooling", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "827a1141138c3d8f4ad131b1f9d8c368cf9ceebb"} +{"id": "train_06316", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5c327ab04242d90cca8afb97dd237d04cb32c71"} +{"id": "train_06317", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d03d3439e27fe0ed5fa67b815b3f4ae00af4e70"} +{"id": "train_06318", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "evaluation_metrics", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ed7fadb6e2ef2e72b00db96ff46729ac9c6cf3a"} +{"id": "train_06319", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ee6772d15cdbd722b80d8c2d84d2aeec4301826"} +{"id": "train_06320", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8e735867df236fb31552954269aeccea3123fa2"} +{"id": "train_06321", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e33d3ae7d3dfb6c281dbf012bf0f4543eeab25d"} +{"id": "train_06322", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "evaluation_metrics", "governance", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e519c5f9aa05b5dd8ac993d2505d86473b1624a9"} +{"id": "train_06323", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6e800f775bbaa306d146075dc57570f71be37d6"} +{"id": "train_06324", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4e3429065e2e69c9941698710fd0b3ec62f16a9"} +{"id": "train_06325", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85dd76db4582e631803d25c05f34b6aa2d72eab7"} +{"id": "train_06326", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9c056568a222e174a1ebe897570b793e6515b62"} +{"id": "train_06327", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8180172c783d4bf8d5254abbcfd1b82d3951ca03"} +{"id": "train_06328", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "ci_integration", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fceae268ce611d1920122539cd25805ad8f2720"} +{"id": "train_06329", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd27d4cd801f7f2624e56829a8edfbad0802aff9"} +{"id": "train_06330", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dddcfcdb9afc4fce0f99be37361a95472d5f565c"} +{"id": "train_06331", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c8032b6f55700f46cb4686a10a10466b574ee31"} +{"id": "train_06332", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "governance", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "651527682d2c7a94bdeb1777098c7c73ab939849"} +{"id": "train_06333", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28d5fe256d9abfd608596e85eff54907c2dd82c0"} +{"id": "train_06334", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7649af5d4f27d370206d7cebe771cba562025ab"} +{"id": "train_06335", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4654cdda96c0b5152ceeae33e816dd194975c0e5"} +{"id": "train_06336", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcafe62777ab4de08e2b41836646228f4ddf592a"} +{"id": "train_06337", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5af587e52e4db9b2d92b41caae633440b6c5ff65"} +{"id": "train_06338", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8df487e9c69504d84b849b016023eacb5a9a267d"} +{"id": "train_06339", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "governance", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f84d98930a0ab4c96bbb0ade0df2a2e24d085fc5"} +{"id": "train_06340", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce77152751de41978276f870ac11f989c181a0eb"} +{"id": "train_06341", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0e216080364cab431a47b77a28bc73f577da022"} +{"id": "train_06342", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a74cad288391bb5ae2656e06cc48b9b9d30a73e5"} +{"id": "train_06343", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "288f29b1bd60e76339565134f7bd1c98f03ed2b2"} +{"id": "train_06344", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "governance", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afe2fb0e021c7dd542ff130482cbc30fdece891f"} +{"id": "train_06345", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "ci_integration", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7deeafb6097c3bb3532dafaa9336412e23e1705"} +{"id": "train_06346", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1eb7d93417db2807303bfa03346affcd6aaa439f"} +{"id": "train_06347", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ea34a72c81cab07160542f0b862224dc5113cfa"} +{"id": "train_06348", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c3b380c52d763017171b5a8b6feb60ce47c90d1"} +{"id": "train_06349", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "auditability", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d184e4635c1c650cb3f959431bc1f87dcfd20f4b"} +{"id": "train_06350", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6eb73c90ea6d49b884bc75eb420ade21717b128b"} +{"id": "train_06351", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b225a5121bc78a14222fbe75854a5bdc0045bb4"} +{"id": "train_06352", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5ed8baa86bbf24cf575eb40a92f1075e94d4a19"} +{"id": "train_06353", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "auditability", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87361510ee0d276453712ae0171d9c640cbb29e9"} +{"id": "train_06354", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80a131a253d4164ba12ebe4da03284a300026617"} +{"id": "train_06355", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "ci_integration", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "927333f72c74d5389ab6d35ca85a21beaaecfa52"} +{"id": "train_06356", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "documentation", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0033e5e147835f5589212bde897faa08d6b74875"} +{"id": "train_06357", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "tooling", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34c03c3c932c618615a0495b5f0eb3794f65afa4"} +{"id": "train_06358", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebf8518e7274de630ebe964a88e8fcd8e6265418"} +{"id": "train_06359", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7771ada393d5fc21d2f45f2496b4953d860e698"} +{"id": "train_06360", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "documentation", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bd73341c693fd9fb6b0945c86544dfc02324a1e"} +{"id": "train_06361", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "governance", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a39fd2e1925349c1dfb2c92b86118f52b8fb6285"} +{"id": "train_06362", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "512b8bd43a126e6758ac4c47fa6e67bc70f25eec"} +{"id": "train_06363", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3eedea4a7d3716d75ef6bcf3d97eb9d0f29d36e"} +{"id": "train_06364", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tests_are_truth", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "932a63bf8fcb16b1dd6163fd3e2a2f5241c3ba7c"} +{"id": "train_06365", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c28ce615447d948bf6233ea2be161037528c0d7"} +{"id": "train_06366", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f29d2aded0b4a722e097e3970830fd3c9f89d1c"} +{"id": "train_06367", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9e3e8da0cfe7c1b6a326628627506f5e355b223"} +{"id": "train_06368", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7012427da08b93070916f45e327d179d47749983"} +{"id": "train_06369", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afee596d471c108ecb0c671d64f8c8ea1154daa5"} +{"id": "train_06370", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30b8178dae4d47af16670bfb7a1ebb3174388f83"} +{"id": "train_06371", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "tooling", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a48587aa0cdcb84dedb6fa049796e73e23290cc0"} +{"id": "train_06372", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73084f94c0dd30903ead82b7444c09513ac474d7"} +{"id": "train_06373", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf4cf8715916b2a5c77a51e338357a1e87954541"} +{"id": "train_06374", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "security_gates", "tooling", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d1b3ba63d41f8753105c8e44644919cb0d0c5b8"} +{"id": "train_06375", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "security_gates", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2708d552ceec24abcb6a87a5afcd198f718b1ad1"} +{"id": "train_06376", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "documentation", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "741ba1025e12de9e423d7081b4661e93f92d5a3e"} +{"id": "train_06377", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "repo_scale_reasoning", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "010c870147d52b2871c600234845b3846fe36844"} +{"id": "train_06378", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28b8f548ef906092b42b7273994a39fcd901c6aa"} +{"id": "train_06379", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "auditability", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e0afc096d69b65b59c08cbf00ccc39511b049e7"} +{"id": "train_06380", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4568a6a944678b12a5cf11e1d270ab29b1edeaaa"} +{"id": "train_06381", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tests_are_truth", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85a5ee0191f5b4c83cc903b3209d4bcf3f9742a5"} +{"id": "train_06382", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fd0a32beb075771a11fa75888c9e7c663c42c86"} +{"id": "train_06383", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bb857eddf2aa3d5490d093fc9616d88b8e2900e"} +{"id": "train_06384", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "repo_scale_reasoning", "auditability", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a67e7a684673f3fe44e570d1baaac60cd0b33276"} +{"id": "train_06385", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bda8e474c3efce994309948d63adc477ab774c92"} +{"id": "train_06386", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ade08aa12426b71b4633cb311e63718b964d5022"} +{"id": "train_06387", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78c4c8f7a285243cf5d43785f7c1fdfa4ce2cf62"} +{"id": "train_06388", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26e9ac969a073ca6be22e8e916beaa4c4449ecca"} +{"id": "train_06389", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4523bab3d6f137c7681a1b8ff1cb00eddfca8878"} +{"id": "train_06390", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "evaluation_metrics", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2edf0d259227290dac65b1c1145a62173bc36fea"} +{"id": "train_06391", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8e84e5d69a67c1d19dd9c4a6326c31f48750948"} +{"id": "train_06392", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "governance", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb00ef44656a444498b5ebd60334c19daa0d50dc"} +{"id": "train_06393", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "auditability", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "edad4ebcf503c58a376b050cd91a5289ffac7437"} +{"id": "train_06394", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "documentation", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f61b7a4510b1d6e6bd0618696af1ab6962266b1"} +{"id": "train_06395", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "ci_integration", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb3006328eae7b2f78bae436f47253e261decaa0"} +{"id": "train_06396", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "365356b0521319767702b3df054731e82ee7292b"} +{"id": "train_06397", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2a479bfc5d1d5845fbd89e46f0c96c283b31d4d"} +{"id": "train_06398", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "governance", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b853b4176fb835af9318085918893caff9b84d2e"} +{"id": "train_06399", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea7e3f5169a4f88ab89f7ab839bb7f6005718e53"} +{"id": "train_06400", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc91147bfcc886e6845c7282b45e475a7862a10d"} +{"id": "train_06401", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f4e232c1f0b8646b352deb9e704380f12478fea"} +{"id": "train_06402", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d5c5626db9b5971c70e5949244e9763ebf4fa2a"} +{"id": "train_06403", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c09a8cf389491c562270d3eef0d335e86ad9143"} +{"id": "train_06404", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d1e8309f7957233ed2b472d8395bd57aa2c667b"} +{"id": "train_06405", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "992a4f2562942634cb5e60fca79075e13054ea00"} +{"id": "train_06406", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fda2bd2ee98ae683ba6ff407324a1ea2a7db0cc"} +{"id": "train_06407", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "reproducibility", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7d99fc793a1c140a82bf6bfdd55631b7da4fbff"} +{"id": "train_06408", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c6bc322dcc34f9c9e696fbed2fe6de10ca5693d"} +{"id": "train_06409", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d87209880e6a121befee6dec490699d104aabe7"} +{"id": "train_06410", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a52d4defef28a8df8c1290ccfb4753290bed862d"} +{"id": "train_06411", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7454c5c1a0e2e1bdd8ef1ea7d42a13128437563d"} +{"id": "train_06412", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a97488409a3194c4ec682209301122ee330d1903"} +{"id": "train_06413", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "repo_scale_reasoning", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41602f7f12b7b37ed5531ae81c486a8484f8948f"} +{"id": "train_06414", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "governance", "security_gates", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da57d50bd4b5761912581fbad32357b021531bcc"} +{"id": "train_06415", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d849270e865d8cf3096eebc788d260df92b0129"} +{"id": "train_06416", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "358ce75b003180bfc4725518d38ef9bf6168d7a5"} +{"id": "train_06417", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aad241e1c9d6820f45ee81e9189cb55679e6710a"} +{"id": "train_06418", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23a221a47dc3d1fd3bb74c4d0fd848da69cd1d7f"} +{"id": "train_06419", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46d5c688b9e41d77a486d1cb940ca7ef1c79fbee"} +{"id": "train_06420", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47f3db544d5c4374b886ff894baf34c4aa1ae988"} +{"id": "train_06421", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "ci_integration", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ae53dc91f9f1336eac6f55678547d5ac1e5193f"} +{"id": "train_06422", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "648230042ff698ded340c930140a9d7025774b6f"} +{"id": "train_06423", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c3fda2c971ea63f8b7718eaecba84b1b039d3c8"} +{"id": "train_06424", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "858c149bb577fe726ba3e029015a2c51daab2d72"} +{"id": "train_06425", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9e7bd26b63a02e479200a63443d628cc8ab7cc4"} +{"id": "train_06426", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "governance", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d0b7de09bb0fd36b6cd5c802ed822a53fe8d9bf"} +{"id": "train_06427", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "tooling", "governance", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc923b44027ca5c527bf5edfa3fc992093d30218"} +{"id": "train_06428", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ede9b8d2f505f57a28bfb719b9a5baaa70b1c96"} +{"id": "train_06429", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "846b7de43f8217e3710fd13cb8b22503728618ae"} +{"id": "train_06430", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e46fc609b779392c3eb621c32e89e294812214b7"} +{"id": "train_06431", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "reproducibility", "auditability", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e27bfc662d0a2db909d484324730703b410b07d"} +{"id": "train_06432", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "tooling", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "587c3810804546943e00c663dfb81ac14947cb9c"} +{"id": "train_06433", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd2c361c94acc4ae5eecb2464b855231b56da070"} +{"id": "train_06434", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67d87a9c5d73519068b5cc9301459e7621eaffab"} +{"id": "train_06435", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba735827a3359d0ef083253102a834b0efaa0e47"} +{"id": "train_06436", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5935622e0f812762e75a52b4d273f52e862938a"} +{"id": "train_06437", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1b5a608575a826aca1e683e534d10042c279d96"} +{"id": "train_06438", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "evaluation_metrics", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce1c12f977f3f3420e858356ccbebccc95d59f49"} +{"id": "train_06439", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ec6bda08136c38d9e74e225ede5112ce7dd556e"} +{"id": "train_06440", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "ci_integration", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3fb5189fe3ccf266b0a665423cac84eab509797"} +{"id": "train_06441", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "security_gates", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1811dd357532584b242d6dd5c6c4625eff4d1376"} +{"id": "train_06442", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "483834f90ffd615cf7a4756e18b61bb784a795a8"} +{"id": "train_06443", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f520e364f5d68ac498fabe6552df1954381faeb"} +{"id": "train_06444", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0bfd17eb71e026e300f72bf8b63dfcab48f4b5f7"} +{"id": "train_06445", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "992a4f2562942634cb5e60fca79075e13054ea00"} +{"id": "train_06446", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d853040a39cf6c626dfe64e965b6050e66c33a6d"} +{"id": "train_06447", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fc0357265032fa6155d1933ef7eecc8a73858df"} +{"id": "train_06448", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3166df9a678c95bfbe651947d3e9ec74865d1516"} +{"id": "train_06449", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "reproducibility", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d9938db535c0f9db711f5046514a96ff2ca3141"} +{"id": "train_06450", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ae8305c3b7919e956f37dbd3ed81dc7f75dccc6"} +{"id": "train_06451", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fd6dba57b7e24fa8a907383c8ff51b7255f3ecd"} +{"id": "train_06452", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9422a0bd579a981d876f568aae15a6e5700ed1ce"} +{"id": "train_06453", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16d363ff757c156e3f93487980f1de1b997f5e54"} +{"id": "train_06454", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "evaluation_metrics", "tests_are_truth", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0df682007ef6a32f480d90e8b889d6231c3e8fb"} +{"id": "train_06455", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0ad47afba12f32b9001ed3eae760aa39ca8dc2c"} +{"id": "train_06456", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa13909170668c2cee69969f015581ebc1ea9333"} +{"id": "train_06457", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5b5e3f570240474befdd66dd60cac6549675ad4"} +{"id": "train_06458", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "384f650b4c0b7cf79bc8201078187fe5b3c061d1"} +{"id": "train_06459", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "reproducibility", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6ac51dee9349a5815c4fba83cd4912a17b6a0e6"} +{"id": "train_06460", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b38f24eb0e04c4f71e105d6d6c7ddff01d57237"} +{"id": "train_06461", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9965357af191ff91c29ec8fcfc509c711f726825"} +{"id": "train_06462", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35e3c373b345a5040c4ca23507271596e144b178"} +{"id": "train_06463", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2aaed0f0d579c4d0d966d7ea21578581fc4e9c2"} +{"id": "train_06464", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tooling", "governance", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fab73c9817fc33e1c2354da686ac5de2b74cbee"} +{"id": "train_06465", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "reproducibility", "tooling", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf84a303061d83155e3f75f9a28bf32d30f38e16"} +{"id": "train_06466", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "auditability", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "689e8dfea625544e3f5066ee02afc6496e097f73"} +{"id": "train_06467", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09f5c6c684ee62d779313411de6d54faaa9673e9"} +{"id": "train_06468", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e656538733914f1326033da33fbbc92ae9247caf"} +{"id": "train_06469", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc24281631605fcb2d99581097edf56e04473c40"} +{"id": "train_06470", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "379eb6b2ba9932c9c10a6da5097bdbb239dbb9d7"} +{"id": "train_06471", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf1bde98889720e6ffca5a34fbbaa8bddcce58f6"} +{"id": "train_06472", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "security_gates", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a68f948b9301ace61336c153a9ad16be9e33ba6d"} +{"id": "train_06473", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "834a78c4fca2f9552aaa89d7b475c0c028a499fc"} +{"id": "train_06474", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "194a585e7237a4d8e5e6e5188aeadd71f272918e"} +{"id": "train_06475", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "reproducibility", "governance", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "550bbaf86ea65e178ee51623bb94c58837461f1d"} +{"id": "train_06476", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b33c4e5533b52d3f5745428140d04eeb8f3ec833"} +{"id": "train_06477", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "ci_integration", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b276c42a8ab549c248a827c040fde938f8c8620"} +{"id": "train_06478", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad363dc07fb084cec9429b75ce0c702286a0c7d5"} +{"id": "train_06479", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e16578a1895e22fbbb8f19e8053eef0ab3145ed4"} +{"id": "train_06480", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "reproducibility", "governance", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fd5266835b2d8216280a423275bb418ef1349a8"} +{"id": "train_06481", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tooling", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa1db1df4d64f97f3ae6709757f75d686de161b4"} +{"id": "train_06482", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "975b43f19a2a489f2cbe6105bc61e3ae7edc0c39"} +{"id": "train_06483", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80a341b502dce43351458d77d8398e97b81836c2"} +{"id": "train_06484", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "ci_integration", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed4dbf94e228f0243f0f90bf01a46f035c5c5dd5"} +{"id": "train_06485", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec5d3967c2057f14425e7258dfc3929d976cf5bd"} +{"id": "train_06486", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f6c7656453f0a74e86f1e41b2711df0255ee140"} +{"id": "train_06487", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "governance", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93e4751fbbd7062753a4ddbf8d003ee728ce5ee6"} +{"id": "train_06488", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36ccf8d160aa32177598b80e43ede6e05e6543cc"} +{"id": "train_06489", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "auditability", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92e90975220349cd88504c4fc7f30efb76f3f1d4"} +{"id": "train_06490", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8d5cf67c9fa01709bbfd7caa14007267bee5c7b"} +{"id": "train_06491", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d279fea0fdb8c5d426381a15007d6b79228b049"} +{"id": "train_06492", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51d2ddabdc4963fcf6f5ddd3ca195cad16fdfe56"} +{"id": "train_06493", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7644151b16a1a42ff7aa44021451dcc0e292f2c"} +{"id": "train_06494", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "governance", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3682cb717e97be48e7853a32f4d93920e62015aa"} +{"id": "train_06495", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "security_gates", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1faa206fc53c97cc165ded062a95ed3e5689f1f"} +{"id": "train_06496", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tooling", "auditability", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d47d571773e2999817bdf1c7c0e67f1449091f0"} +{"id": "train_06497", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "auditability", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5907d9f635fbe071a5a53a706aa0f0e57331b131"} +{"id": "train_06498", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "243893d38ffcee49347a8e012b134f3e63b388b5"} +{"id": "train_06499", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bbdd9dd2d43f45d646cdd431b2d24af8eddbf8f"} +{"id": "train_06500", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "security_gates", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7de2f80ded34c23528a585f09311dfb040a5075f"} +{"id": "train_06501", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97f979f9517da9b062e1eacf2fb865701573ca7f"} +{"id": "train_06502", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a48587aa0cdcb84dedb6fa049796e73e23290cc0"} +{"id": "train_06503", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "ci_integration", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e281a8308799d381215fa2a733e7c1810dd9b67d"} +{"id": "train_06504", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "461e2d5858dfde19b736653b36abf8bb7d7d335b"} +{"id": "train_06505", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f42470cf04a6f466119514a7074d77e305e83ef7"} +{"id": "train_06506", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c21565a76dea538e7383e54bbd486a0a35f5b9b"} +{"id": "train_06507", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae5877110e6c9df710ae022e1a460b797ed607f8"} +{"id": "train_06508", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "evaluation_metrics", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61975afe620be03265d3b5f18a642296f7007f58"} +{"id": "train_06509", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "auditability", "governance", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5cf58fb812017b7119247fe721a1979889c4b4d"} +{"id": "train_06510", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "reproducibility", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc0fd7f3e9e728f6103b2630ecc0587ef79a9d6d"} +{"id": "train_06511", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da2e27e432c145b661d209768475734294d92470"} +{"id": "train_06512", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d0cf47461f0db051f9ff5f76d164c1c2b82f0e0"} +{"id": "train_06513", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "auditability", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea1cf397984a67b7e8c19b38836236fb0831b6b5"} +{"id": "train_06514", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5e3668e02ea243df5d9214fbe4d782fcd6064f2"} +{"id": "train_06515", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "tests_are_truth", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3fe9149b4d03fd11c928d32578059dc18cfab90"} +{"id": "train_06516", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f05a3da4f0ecd0be6971bddb5076b89d1a4cc65"} +{"id": "train_06517", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ba9a1d300bed3758be39b89b9b2312c343b516e"} +{"id": "train_06518", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbe7c45eefe36ad61f857cd08bb574394c1724c0"} +{"id": "train_06519", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "ci_integration", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0690c7c6aba656b7c6316df3ac973a5ad3fb1ec3"} +{"id": "train_06520", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f1e586f9bdcacefdc4aceab4783f69408555d6b"} +{"id": "train_06521", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "reproducibility", "auditability", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b807afc58f3527d9939e2996a553a8c7766c2db0"} +{"id": "train_06522", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ef6064becc4967c0a4fa48c848d298afcfa5123"} +{"id": "train_06523", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dbb3106fa4b67d40733258302a2016ada6c3ba7b"} +{"id": "train_06524", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "492c12e8a64e2de8ea4c56ced652485430c01f9f"} +{"id": "train_06525", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "tests_are_truth", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "700fd55535058ced437597a7d12c17ac86f31c6e"} +{"id": "train_06526", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "security_gates", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d615f811619b45e741bf28f39c77d1812fd49541"} +{"id": "train_06527", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tooling", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d922e11308c7480fd2c0e5bf18f7558f634063ed"} +{"id": "train_06528", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21247bf6ec5b436158de9047059e8f59fb2c5d80"} +{"id": "train_06529", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53bfaca298ac755cafce0b158551a6ab8c41c191"} +{"id": "train_06530", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "558c308f0ce2dc4e77515daefc75bab2cc3c6cf0"} +{"id": "train_06531", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "238c7951f73651347a2ddb41b1172e04e4e7c55f"} +{"id": "train_06532", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "reproducibility", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64d401851a2700741a7c4bbff31382b8c3ed6146"} +{"id": "train_06533", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "599c219a983fd2581537bb489b916c5147c560e4"} +{"id": "train_06534", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "103ebe6af27e836c9bfa8aecb0f0e37815e336ea"} +{"id": "train_06535", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8085fe2b613a6702a8db8ac7f3ad0fb2fc3f493"} +{"id": "train_06536", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "ci_integration", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6c024ed744735eb47e0a9aa8d821affe2a1c05a"} +{"id": "train_06537", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2ac29bf14d8921a176f2e264947045abf4b1ccd"} +{"id": "train_06538", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "governance", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f99eb9a9228e4d463413e91e7bacb903653d33f"} +{"id": "train_06539", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0aefba1b07938e0fcf22acdfd256283306479aae"} +{"id": "train_06540", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "documentation", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d189466cb4029e394c33a666b184c828705c2c4b"} +{"id": "train_06541", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bccb6f534aeef2457ac0388ef499f68b289c470"} +{"id": "train_06542", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "ci_integration", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13fd302eef0c3b838180f7c954249cb01e9a8dec"} +{"id": "train_06543", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b91d2835562fda523e5ae905ece4dee214e5f5a1"} +{"id": "train_06544", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "evaluation_metrics", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82e2c922152706ec553df092e302ed8ee3c6b578"} +{"id": "train_06545", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f095d746bcf6695a1b998c1db9654c74c9ff9139"} +{"id": "train_06546", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a46d829b66b509781723ce743b5caeb0d6506fa"} +{"id": "train_06547", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "134bde94580531c1e8cab356890a9d916c4b42d3"} +{"id": "train_06548", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3a8fab7bca6eeaa6a729c709471146e7592a29b"} +{"id": "train_06549", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cef8b51bc01e85e24c08ed89871c164a5540578"} +{"id": "train_06550", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "governance", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a33166b6a2f246fc6ed666bc689150df2e6a2d07"} +{"id": "train_06551", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab0ea050403be478d5d2119bc0e53a735c8b8c34"} +{"id": "train_06552", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "985835b9947e453983b92422d6eca8e43d6b1d29"} +{"id": "train_06553", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a937e28bb702d449219858b469316b51bb0c778b"} +{"id": "train_06554", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac09899250bde01c7fb4236f186c314529822c7a"} +{"id": "train_06555", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88e6e5c95bc11e8586d4afd7a702f3b380e47d04"} +{"id": "train_06556", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fe6b44f28de1008e8b3cda7a7b3595d225f81bd"} +{"id": "train_06557", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37ca5d1c68ef6bbf6f412da295712fcaff1a7fe8"} +{"id": "train_06558", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "143c77744f277af075534398f7da2065444f9ed8"} +{"id": "train_06559", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb2a9301b8ac14b50ea16fbf2a0837c6f65998b9"} +{"id": "train_06560", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "554a165ebd8fd2e44f095a041e8e77beea5df4ae"} +{"id": "train_06561", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "reproducibility", "governance", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "743e2bf74b7e14981e02382fb135c09c48db1a35"} +{"id": "train_06562", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "599693a97a4d3bcdf8aea33b6a5f064fef60be05"} +{"id": "train_06563", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "tooling", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82719cc96e389770ac27ea054a22ee5d9e73774d"} +{"id": "train_06564", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03ef7abe2fe5dd17ee94d77e61e177d3220594a5"} +{"id": "train_06565", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff76105755eaa066094c491bc87cbe82a7b987bd"} +{"id": "train_06566", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "governance", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf79aa361c239ab94768f67979d9a6c92d1f2a92"} +{"id": "train_06567", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8870f863d614d2b60258e713449cf6ffba909b7"} +{"id": "train_06568", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "governance", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "813619a79aaf18ff0d10e521e5f19965c5b77e44"} +{"id": "train_06569", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "auditability", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c047e725551bb29efc77cd10c918b7d1b48f6ca"} +{"id": "train_06570", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "governance", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cb3fd1a71ec7bc2a28d8a5c5fd527edae44cefd"} +{"id": "train_06571", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "tooling", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1634a5fa711995dec0fa6d7f1eb8c9d88de2f01"} +{"id": "train_06572", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fbc935c9a8632f3d657d840b3f626eca9f16c5d6"} +{"id": "train_06573", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cd01863507627a8ae1f1f1a0eec9f07f9a1a386"} +{"id": "train_06574", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b61078a9336a81f3caac3e6479f873e2b8a79547"} +{"id": "train_06575", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eab87991636e04572dbe4fddbc6c6e10c409ccb5"} +{"id": "train_06576", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89d3b7d81166d58920759b8c5a0b0b38d96b37a9"} +{"id": "train_06577", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cab07b1db1382b0a1b75c5a9517db92e3b5fc59"} +{"id": "train_06578", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "auditability", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07aa946516de48383f3146c1d8125abfab1d6a76"} +{"id": "train_06579", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7ff266eb60b027d67395bbab8b0e7632e6757ac"} +{"id": "train_06580", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f80196915192b409205d8dd63f83d40a641e9b2"} +{"id": "train_06581", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0c47d45d15ec40775403edf6e899b088aab687e"} +{"id": "train_06582", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5e3edd3267edeef928d032dd3110e3976531e8e"} +{"id": "train_06583", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "tooling", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9dbaf3ee9572b3f2e9d33453c0e7df3a3923af1"} +{"id": "train_06584", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d719ef5974cc250fc351b1f583921a6a82581fc"} +{"id": "train_06585", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61a7fba469fa0ca8aae865df765f5d3b673d08e3"} +{"id": "train_06586", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6fe28c6b240180a69666b904d31812548e52a0b"} +{"id": "train_06587", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c37269eff73bfde3fb410addb308b6084f46b397"} +{"id": "train_06588", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bac47152894583a8a9dd7c9b91a510fb621ea5fa"} +{"id": "train_06589", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tooling", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5b836cef69f521b42643d5acf9fcba4e8faa067"} +{"id": "train_06590", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3c3d5468f95791adf22119ba56669c4a140bccb"} +{"id": "train_06591", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c88a1545522e11c2af630c310a1e92450341709a"} +{"id": "train_06592", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f24e9fac3822a5995730dbc4107533f83ab1dc3"} +{"id": "train_06593", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "tooling", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "694609446e654e645d5e458752231ebb599c588c"} +{"id": "train_06594", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "governance", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9e393646230998889595632c8d516ef1fa294f9"} +{"id": "train_06595", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2640a467ed5cc1a96d65b138e7d7df83b847344a"} +{"id": "train_06596", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "evaluation_metrics", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13c24bf7557a0d2b5ba2d48edd4249d135d424eb"} +{"id": "train_06597", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "593839ce9172bb42f27d7d4662f7c9cf73f6cf2e"} +{"id": "train_06598", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tooling", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a531ddb5932470d5925dc6cbdb809fb3b208199"} +{"id": "train_06599", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f791b86c1f586bca22ed96caca709e1251525aa9"} +{"id": "train_06600", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01f858482c732721a1d6bff9ae4dcf9cfa6e2d3c"} +{"id": "train_06601", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d249854c3351920e816cbd8ddeba1c2315932c1"} +{"id": "train_06602", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0b95fdbb49c8d5fa3211b5dcfc5056a7a243a36"} +{"id": "train_06603", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "reproducibility", "auditability", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66f61bcf44e0a6308b0b849d61fb1ec71b2e568e"} +{"id": "train_06604", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b37796be848a699bac9650db57bb3f2f9e095c4"} +{"id": "train_06605", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "governance", "tooling", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3eb0a585bdf23ada423d04e42c7155bdbb725592"} +{"id": "train_06606", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccdc9fd397e9189af5524c891ad92a12b04e59dc"} +{"id": "train_06607", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3e662a4977cb1451c61d4cc3041e464ae58b210"} +{"id": "train_06608", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f39952b259cc06aa5dab2d59b64c489efba089c9"} +{"id": "train_06609", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "security_gates", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3e0d36bc2d25026177bbfc402521a082a702656"} +{"id": "train_06610", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "013580cf1e6ec798d7eb5020fc54d90d6f39938f"} +{"id": "train_06611", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23da9fb2a25140cbf4b1e9650f857a8aef865eba"} +{"id": "train_06612", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "auditability", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1eb59525d087f4a83ef1cd6a1d4d1a49ec0d631b"} +{"id": "train_06613", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33bfac3a5ab8eeb63118db2ff11b1f1d41b5627d"} +{"id": "train_06614", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7565003979201d4c31a2fc4c918231fc73c2409"} +{"id": "train_06615", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a66a2bb58f551fde26b0a63b4791bcbd98c989ce"} +{"id": "train_06616", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a0f12828f5bd646d47b9a400b12f3bfbb73a98f"} +{"id": "train_06617", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "documentation", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4e93d78c5e3278433b518ba7691b4fe9acff065"} +{"id": "train_06618", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "ci_integration", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3df60378e76127a9ecd773eecc575d50bfaea328"} +{"id": "train_06619", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f92b714d9f570f9ee46b664f0d4c5baccd88ab52"} +{"id": "train_06620", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "documentation", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80509e66eaa7c0af3a1ca7508d2fc24f22d0d43c"} +{"id": "train_06621", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f1e586f9bdcacefdc4aceab4783f69408555d6b"} +{"id": "train_06622", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52661435d684c2ba7f79c447ebc87a0d11f8926c"} +{"id": "train_06623", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6208c8cb7a16ffd14db569633567d5d13e08ad7a"} +{"id": "train_06624", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4123bc2c612e3010b127d6882c34528f46fce71a"} +{"id": "train_06625", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0bc40701a8450f1b0d8eee458b22a930aa4f112"} +{"id": "train_06626", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4dc146e42fc87d8a6a4c8263673bdf9967422f1"} +{"id": "train_06627", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c911c35f6adb480509adab50926a6f35964e35e0"} +{"id": "train_06628", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40b8349662050a3cd7953f6f88b34d1a2d6c0355"} +{"id": "train_06629", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fc3ff6875a64b61d8b90a4cf0abc8e28e2ad921"} +{"id": "train_06630", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f190cd0d9c1359f868e81f1a928c807c1783fbe"} +{"id": "train_06631", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c25b331a199474ae8631a1a97eb7d8ec13a34cef"} +{"id": "train_06632", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "repo_scale_reasoning", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc682a53d95dc32aee5ce70b017aad3fe88a7203"} +{"id": "train_06633", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c09a8cf389491c562270d3eef0d335e86ad9143"} +{"id": "train_06634", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d939d9f766d1892a95d0b6f078168dbaec84c2b4"} +{"id": "train_06635", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a2b49d7d04c4f2b0a7842e1d4cbefb167414090"} +{"id": "train_06636", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "cost_latency_tradeoffs", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54b5a7e829d80699d8e2b37667a8adb4aa17a36b"} +{"id": "train_06637", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3768d71cd78f2e0cd11b4c8f85d513cf75b45017"} +{"id": "train_06638", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "725fa9b753ef8aa37fcd3611a7d6c33d56d41540"} +{"id": "train_06639", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "618e817e1851541678f726c55829a94f0cf04ddb"} +{"id": "train_06640", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "050718c114a687f42dd02bd598675fdafb7ac33d"} +{"id": "train_06641", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d26c4399a52eb6878cfd2ee399d9695365c7535"} +{"id": "train_06642", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "documentation", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70f4af87dfde7b9a32db19f0b8e07d7fdf98bc46"} +{"id": "train_06643", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "evaluation_metrics", "repo_scale_reasoning", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a47e1e0f5d79da33ebdf78100e552de9db382205"} +{"id": "train_06644", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfc67273cdf15e87aa9c0f8c09559c8579721c3b"} +{"id": "train_06645", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f327d6a48d9c824019836907f7a4a21b50ee7710"} +{"id": "train_06646", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "600f35a9d938a785358459ae24839e935936ecc8"} +{"id": "train_06647", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c70a6f44075c57b3df829065a346f7de60872c24"} +{"id": "train_06648", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be4caf35cfc291740d25276494a49d43894f55ca"} +{"id": "train_06649", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02f84503e746944627b680c53e59a15700f70018"} +{"id": "train_06650", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b13dfec876a4f30840bde516386dc6f4dbdbdef"} +{"id": "train_06651", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19593c1195bf41473cdcf60bca41e20869a6784b"} +{"id": "train_06652", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66833561885674b1503ef773f048f8f15f067bc8"} +{"id": "train_06653", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "320b5b0455526ae493e9419f3b637e8b290385db"} +{"id": "train_06654", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "tests_are_truth", "governance", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecab2ed9410362ac4ee10f3a5f81a180d95b318d"} +{"id": "train_06655", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac123c4b7b3e4d6cb82d133ece0bd199d778aaa6"} +{"id": "train_06656", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "governance", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae9b03d45fcd01a7adc1b3675f2b33ab3b0f1c0f"} +{"id": "train_06657", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "817cf340ff825502c122a2be45913aed7aca1cc5"} +{"id": "train_06658", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70ffcd38ba12e32506d737779bed077655b8d631"} +{"id": "train_06659", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3990fe06732a85c06c5ec65ad78d93d57b1e37ff"} +{"id": "train_06660", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "716244f33ad0a93fb99c33cdca8b1bc9e767ec34"} +{"id": "train_06661", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ad4286e6a79a849275579fae3fd5741a49d89f9"} +{"id": "train_06662", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "cost_latency_tradeoffs", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7f191d25046b1fb00634d519038e012cc7fedde"} +{"id": "train_06663", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "security_gates", "governance", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb1ddd399277687f9948d266fafdf9e2ae3a74bb"} +{"id": "train_06664", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "tests_are_truth", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0eee33d8eb79272c09e16f9559b0daaa1cf9f007"} +{"id": "train_06665", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fae4bdc2b725d31224557a1c3b6ca81d83e175b"} +{"id": "train_06666", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b33f34c9c8b5f88b1e477053800592ef4bfa7b7"} +{"id": "train_06667", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3c3e4390eac9c727a7e5f099357c23d1f346656"} +{"id": "train_06668", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7406a606fa32fcacefa8d9e79843b2301b12786a"} +{"id": "train_06669", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d06dba29cc434110560eb6549ec834777af299af"} +{"id": "train_06670", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "tooling", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f2d2a75c861e61af211b2f132f68cae1eb06441"} +{"id": "train_06671", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "security_gates", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5cbab58a1136d15625ffee8d5b5ba3339ccfcf45"} +{"id": "train_06672", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tooling", "governance", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7ee8cdde2f0e9d9008e32379251aa62fafb1332"} +{"id": "train_06673", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f2bd183c4ca8925bd4315f05d72d87f9ca9cd1c"} +{"id": "train_06674", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "tests_are_truth", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "985aff40728fac272cb3c476278c1b786c793194"} +{"id": "train_06675", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2316773b59921a6b4f068bcafea4895e2b49b59b"} +{"id": "train_06676", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "documentation", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c5990981bde0ee94fad044bfe803c71a8f4785e"} +{"id": "train_06677", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7248a348e7c624e4dae08d2ac1bbde8f7379c27"} +{"id": "train_06678", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0215e77a5cd5015376203717084857944507021"} +{"id": "train_06679", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "security_gates", "governance", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff327afe860cd5cb7dc8f1557cd537456c88b952"} +{"id": "train_06680", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "security_gates", "auditability", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6728f6fbfbf70ab567569604f118371cdaca08c3"} +{"id": "train_06681", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "tooling", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8ec6c73e90995a0922ec1fae6e6cfe6349ecdae"} +{"id": "train_06682", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6744f99415ffbc33be867e82619536ec3e62bc51"} +{"id": "train_06683", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f707c4950b379f868243e39921c5c7aef9e2a89"} +{"id": "train_06684", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "governance", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e14af08fc4ddf48398753d9ad17c34007f720d8b"} +{"id": "train_06685", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "documentation", "ci_integration", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c39fd621aa1d47333641d702edaee3f5b1d030f2"} +{"id": "train_06686", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdef162ccb1bd2090170ef20ac980740ed6a9fcb"} +{"id": "train_06687", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78ed9aafa7c584a48e4f1ffeb97a3819b36d9e74"} +{"id": "train_06688", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "documentation", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b66b790245d10c0d463e7260e2426ad8ffdc3b5"} +{"id": "train_06689", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "965780648d89ab462b2664f9209e51d9bf9dc194"} +{"id": "train_06690", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "ci_integration", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eef057c5d9a2af9b8a7a2261d20c1974ae010ed5"} +{"id": "train_06691", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6dd67450c48ee6179a3d02ec3b2d1a6964b2862"} +{"id": "train_06692", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01c22da15e74ba1a072aeb079125346503c3c1a4"} +{"id": "train_06693", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "788c1d3d763e3201bfdaf191d6fef9364e9bbace"} +{"id": "train_06694", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0df9c4aab35f71573e570319036f65518fcb3ddb"} +{"id": "train_06695", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "969b47d187e87cf0e2e1786d158ec4034523d8b2"} +{"id": "train_06696", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "628156ae0be9347d2e8d664e126abd3870767a7f"} +{"id": "train_06697", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "repo_scale_reasoning", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74d9ee14df871cf632841b6df5ea0c0dda0a8a48"} +{"id": "train_06698", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5aeacd2bbe23905fc594352b04849c173faac668"} +{"id": "train_06699", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "documentation", "security_gates", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e4f932ef20ac73f339088dce7b7863f4df0781f"} +{"id": "train_06700", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6dd0fb14faf7cb13eef1f64be75d65ea68e7db2a"} +{"id": "train_06701", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tooling", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9357cddafc7283e21cab64e4c0d4e7f2423506c"} +{"id": "train_06702", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "tests_are_truth", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b925935383f50393fe44963b12510019d0538d65"} +{"id": "train_06703", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "governance", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90425b9bb2887d3df59e98d345d3fa06bd3599ac"} +{"id": "train_06704", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e0f35071e988f91cfb5201d9c15e87ae4832c25"} +{"id": "train_06705", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5ecebac2cab0654cf9eb6f492d4949f654ec599"} +{"id": "train_06706", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f533b4e2ece23d0f9fba7e69a32f0da0bc357746"} +{"id": "train_06707", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4ba1603754dae4763f9fd73758fae2ea1ae0d15"} +{"id": "train_06708", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7603927e5a2ffe1c7fa3c65abc2b76f04acfc50"} +{"id": "train_06709", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "auditability", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b501f3dc2fbba7b0fae50a7f638fc63ac81c3bdf"} +{"id": "train_06710", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25a2c563bb0fc6cd80520e15b6f342fdce8334c3"} +{"id": "train_06711", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tooling", "auditability", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "32e13b46328c4dfcfdd6a26e4f2915d1e8846b0b"} +{"id": "train_06712", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "repo_scale_reasoning", "documentation", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7dd21ba110235ab0d8aa6283c54c9ac7707b6bb"} +{"id": "train_06713", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "security_gates", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30efd5b79bd76b2957ab26639b2147e36badd8ca"} +{"id": "train_06714", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55e7003b115a0acf58dec68f75d1d67a929894e9"} +{"id": "train_06715", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "ci_integration", "documentation", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00a6b328123a625d01bc9d94ca12c44bcfb3086e"} +{"id": "train_06716", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b239e94dc1c2bc462325b01b6a26166cc827c69"} +{"id": "train_06717", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c844924d8587ce8872f82acc169b876241c921f5"} +{"id": "train_06718", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "repo_scale_reasoning", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "722c2d6a88f933f47fa5bd9834d684015d654e9c"} +{"id": "train_06719", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38e75bec6fceace823d7d0acf78cf8272d0384de"} +{"id": "train_06720", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a543bfd6d0d6b3d90630ec8c41a8a2c9f4f37746"} +{"id": "train_06721", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2b4b7d0ee9863a592c8978f3f8715dc1c7cb67d"} +{"id": "train_06722", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5048f2074d0ba9fbe6a4d5d88d4299cfcc50c8d"} +{"id": "train_06723", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tests_are_truth", "tooling", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b970e5694ac346f3af910b118a6b8a77aa4db3f0"} +{"id": "train_06724", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "auditability", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e7ac289ee1cfb5830ae554724a17235da2ca5a2"} +{"id": "train_06725", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7ca010f0dcf04c39d0b6eb8be18d3bb1cf5260e"} +{"id": "train_06726", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "651527682d2c7a94bdeb1777098c7c73ab939849"} +{"id": "train_06727", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "security_gates", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee95e4657b38ec4334000acab6e4698a9483d348"} +{"id": "train_06728", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14f34453d173ca60a3e6e5b9edb584f2c35ecaf1"} +{"id": "train_06729", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "governance", "documentation", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c166af13aa6c0a95805a0637fc1c1817460b7380"} +{"id": "train_06730", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "015c0d59638430efc626ded4a7f643998fd73a81"} +{"id": "train_06731", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de9c6d193bdd6de106c62234b37bda3c565fa005"} +{"id": "train_06732", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "reproducibility", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5fd55962ba6db5fb769d4369da27cfd2687de97d"} +{"id": "train_06733", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "documentation", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5e12a88daa257ff5cedf440068bfb3c52c5dd1c"} +{"id": "train_06734", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddca09243334266497561e55fb424bceb4a5be64"} +{"id": "train_06735", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "repo_scale_reasoning", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2162faf0af8c856bdd6d4bce5e7fec84fdc1f881"} +{"id": "train_06736", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "governance", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c0cdcbac8d0e20fc6348af696e674096cc692cd"} +{"id": "train_06737", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "auditability", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b639c5452a90e7b11e47b50989f186eb7db8f68a"} +{"id": "train_06738", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f32cdf3cd5581051665e4062fd1d41dc9aed2c6"} +{"id": "train_06739", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6658059ad1120d172bff403fa825f3aa5fc75931"} +{"id": "train_06740", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tooling", "auditability", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61fce0b2cabf47c28c49ff20bcfb05e9e1947c1f"} +{"id": "train_06741", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8781b447f9653c6b631846a41340c407be76a7ea"} +{"id": "train_06742", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0494692566269aa93554b35ab32160c99769bbbe"} +{"id": "train_06743", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "576f549bd241fdea166c8b91d68adef41d11c756"} +{"id": "train_06744", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "517dbab6e178c3290a7fc429109a1919253f48fe"} +{"id": "train_06745", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c99727047824fccbc4f46fcc53013475a2634c4"} +{"id": "train_06746", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6998a19a2ea7c54fb040966bd9c7b8b55ae0f0e5"} +{"id": "train_06747", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "governance", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f9e499609ea62d930ae2768d359546eaa77d711"} +{"id": "train_06748", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "governance", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcd98baef534d7b653859fbf295a75b9e6113189"} +{"id": "train_06749", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "documentation", "tooling", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c879fd673de7a13bd0217b8e61fb119d9ddceaf7"} +{"id": "train_06750", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "tooling", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e93fc0e9b0b17daddcd12a63e4429db5bf7043e"} +{"id": "train_06751", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "governance", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad9cfb5fb1a8b5351db80d9017c06ddd5062fe82"} +{"id": "train_06752", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70fffef3fe1e848984682ee96998bbfcc46e4539"} +{"id": "train_06753", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f33d4d7d3c0d178787dc88a3e0646650c76c5199"} +{"id": "train_06754", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5c91d672cb6e7f077d012d1a753b8a8b82ada5d"} +{"id": "train_06755", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22367942ebd91115ff1df5cda1a3c1af6f57e7f1"} +{"id": "train_06756", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57ea524050525cb2488a07af3e10a1eaef9418ce"} +{"id": "train_06757", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f28856d8d94d49e3715eb39067c35ac4645e55a6"} +{"id": "train_06758", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4452b4e4902f68f34c66b7dc8866b10d767ff553"} +{"id": "train_06759", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "313b50d1a34199cae2af08f1bb58f143e7eeff6d"} +{"id": "train_06760", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "governance", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f4eb0e5d60f2e925a968d9eb25d7efa11405984"} +{"id": "train_06761", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "075299a03394a92f53b340537081191de8ce17b0"} +{"id": "train_06762", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d92763bc72b91073f51052543f40176ef6deb0a2"} +{"id": "train_06763", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d881b93b9c6bdf00bb58eb39918ecc368f4c8b95"} +{"id": "train_06764", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9727471b250d74ca27e0add290b4798b23095b92"} +{"id": "train_06765", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37b8f483d07cbb208a862388ad22e7aad2843f92"} +{"id": "train_06766", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af22c2d6e7aa25920a6cc9e913adcf891efecbf2"} +{"id": "train_06767", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f19c5041d1de88692c48cd32ca50e74a2a9ef3dd"} +{"id": "train_06768", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81ccac023bad1d01c0365ec4937a32e3bb4c7551"} +{"id": "train_06769", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "governance", "security_gates", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d599f21f05562a0154c54f1add15d8f527ae2056"} +{"id": "train_06770", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5512608cb18fbdadb3f74d9fba041dd867f5d7a8"} +{"id": "train_06771", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d796a54b0f4ccda72c637baad7b2b902c66e94f1"} +{"id": "train_06772", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72516350faf61f3db0e10008aafd1063c654017f"} +{"id": "train_06773", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5e5d645d2b73be8497b84144fe95fde4baa5ad6"} +{"id": "train_06774", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17a00d77213653cf9648f10a4ed8d5d52784efc4"} +{"id": "train_06775", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e6ddc2382e36eac0c55decd4d4e52fa35fc76de"} +{"id": "train_06776", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7428259bbfd54a068ecc5ab6b5deb1cec3c43027"} +{"id": "train_06777", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "589cea79f06b30bf029e112cb4d7d27d77c646c4"} +{"id": "train_06778", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "tests_are_truth", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb52feac4a908e7a7bf9e06dd085f7df651f262d"} +{"id": "train_06779", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b8ca3694f100ec11ad7ab964f74cc3b94fd30d8"} +{"id": "train_06780", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e17aba0001283f1a8731a00cabad4cd7fcd0a65"} +{"id": "train_06781", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c00157a1bd597860754197e31366c3f8e4243c50"} +{"id": "train_06782", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afda002eb12abb36cf6ed25c332f88d6252463af"} +{"id": "train_06783", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "cost_latency_tradeoffs", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1e62808a0d5c8f0644c524d4d580836893d30fe"} +{"id": "train_06784", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e61bdd3f80f9c92dbf18d6b7bade4761f2c05dfb"} +{"id": "train_06785", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "repo_scale_reasoning", "governance", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58a22c6eefa11e20fb65c869ed34923a8f3ad27d"} +{"id": "train_06786", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0df682007ef6a32f480d90e8b889d6231c3e8fb"} +{"id": "train_06787", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d46ba87ae0676491a760bf9d2dab0b04bd3389d9"} +{"id": "train_06788", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b94e169b3e1a358b35f3a12d91174a9b5be1ea35"} +{"id": "train_06789", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09bd3afff29dd4fd083a13a211e23a0db71974b3"} +{"id": "train_06790", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e56c8e50f1d02b310d800fe6f752eb8a22592142"} +{"id": "train_06791", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4884693bf1b551f50b19b25ebd21d2c1e3efdce6"} +{"id": "train_06792", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "tooling", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77a7b0199de59e225fc95508be308de215dc213a"} +{"id": "train_06793", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "auditability", "tooling", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "176f48651a1b27835a5ef9c55aa4bef5a2c0f638"} +{"id": "train_06794", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "documentation", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02e38b6f9ae1d3a5776acb64a55f10e417214b15"} +{"id": "train_06795", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f82276d99ae592fb16addd9df7c88f70ef1db2ae"} +{"id": "train_06796", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24a9076a7cbbe5f7f892f4e0a67a1be37b1f569f"} +{"id": "train_06797", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "434c56f8f2ba0f27c4c4b0d56f205927dde939c2"} +{"id": "train_06798", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f64e609e26217eec6ce80212278f5e440336cb6"} +{"id": "train_06799", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b62db0dde5097303aa328562c9a378bae0bf2480"} +{"id": "train_06800", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e721f0bf7688ebcd332d68652ef1c89db1010353"} +{"id": "train_06801", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tooling", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f0b486ce2ff684874510623818d54dd1feb75cf"} +{"id": "train_06802", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40440e7e3963e3af07546674ca8aa4799a5dedb3"} +{"id": "train_06803", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ff00aa2ad098c4adf4cf1107399b6b70e7e7fc"} +{"id": "train_06804", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "security_gates", "auditability", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "220523dabdfc3fca58a4ef150887034a6a42506e"} +{"id": "train_06805", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa60b8205d1046cc8ea28b40da21aa3e9a647792"} +{"id": "train_06806", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7250651ac0676079cb424d8e9cd19535bd1b2fc2"} +{"id": "train_06807", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72fdf8e759821717794bdc272f53d5e828f641d5"} +{"id": "train_06808", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "cost_latency_tradeoffs", "auditability", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4bf30987a0da1d737b1a306ae57a510e3303c3d6"} +{"id": "train_06809", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75badb5e7d0b5bf6bf9296f17f0aebb2c7bae333"} +{"id": "train_06810", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6af1ab25b117bfcd75ca1997b0390ecbd62b46a5"} +{"id": "train_06811", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c749f9e47e25a1796ccf12e42e5c155a5f3daa0b"} +{"id": "train_06812", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "tooling", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9d676d22ddb713671adbc32d299dc6ff319205a"} +{"id": "train_06813", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "auditability", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d881b93b9c6bdf00bb58eb39918ecc368f4c8b95"} +{"id": "train_06814", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1acf084cffc6f02e45768e819b9a04640188336"} +{"id": "train_06815", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cffca0f7e75c59824856f4aa2725a5df41ad8ba"} +{"id": "train_06816", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0cc5001e8c400b021936e2ae4d16aa5f64819c6"} +{"id": "train_06817", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9330186d3aa86af46b95c560567a63b531c7625d"} +{"id": "train_06818", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "cost_latency_tradeoffs", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc76bfd73d0798278e1b76b4f593706f6fc64103"} +{"id": "train_06819", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "ci_integration", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df28adc39b7d1c62300e3366b326cf7bfce09f3a"} +{"id": "train_06820", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "tests_are_truth", "documentation", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08ccc37c9abad4ba5dff485e28d1f0e7d4b37716"} +{"id": "train_06821", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "455c9e01cec4ee6d408fb456874437484cbc0ee5"} +{"id": "train_06822", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a63cdd152c163f5905a8258f3889e176889332a"} +{"id": "train_06823", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3b8f4eb0aee6de5be894fd9771d041cd9375cda"} +{"id": "train_06824", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "documentation", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7f536c8d99c8891b546a248bfe540e96af8a8e4"} +{"id": "train_06825", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "455e34e0adfb8e116bdceb4e1bfac8b7c490a801"} +{"id": "train_06826", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "reproducibility", "tooling", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e44e8e5280e067d801d7126035271cad15d0afbf"} +{"id": "train_06827", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb781d69c15aa4a33a903a45df7c8214ea2efb15"} +{"id": "train_06828", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "documentation", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0978f0bef8473adab6a2636fbc719148bad6b960"} +{"id": "train_06829", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da10425d620a49696ec7d819a2997591ff1083ff"} +{"id": "train_06830", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tests_are_truth", "tooling", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f2489f3784291151dfb2738b7b5f409a51a05b7"} +{"id": "train_06831", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfff49e23efdbb601e6d5146b152ad05b989d84a"} +{"id": "train_06832", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f82b4a683ebbf15b09cf8d20104c4b7d55c9646"} +{"id": "train_06833", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "ci_integration", "tooling", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5452a28d34c917b7e1f47604f897a8e16ff5b114"} +{"id": "train_06834", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "518c73c3e1f23c2a9df99e7dc2924cc2eedf25bf"} +{"id": "train_06835", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22c9eb2ad44997a3807cdf0da1904bdd4e7b63a8"} +{"id": "train_06836", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0605fbc7b313934d6261e15211f12d063afc70b3"} +{"id": "train_06837", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f04e0f6205d42a2dac1103baf7451939fbd45877"} +{"id": "train_06838", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4c493f93a3dfcf26928835f77523507c09c5591"} +{"id": "train_06839", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "security_gates", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3c5e4fa45b2de53e101a0c95c53cde891554125"} +{"id": "train_06840", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "adaacdb406c7b90253e4d477ebdd7eceac219de3"} +{"id": "train_06841", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "repo_scale_reasoning", "auditability", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25e222563278a6732c7a1333c6186cbbb79beb91"} +{"id": "train_06842", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tooling", "ci_integration", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0238ec9c795d82925ef8bcfca5442d5d9ac437a"} +{"id": "train_06843", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1753ad5d5602be652ee03386292828a85c7b19cd"} +{"id": "train_06844", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42ca21e67b76fb6d442af95ce756ad1f33bed0c8"} +{"id": "train_06845", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59d974c5117a9486a7345296f7759fc19a6c32ed"} +{"id": "train_06846", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b018c33f75a78bd3bd9b5d7431807e3af652643c"} +{"id": "train_06847", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "evaluation_metrics", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "159fb99836f89d07a168101208a9e89acfa08830"} +{"id": "train_06848", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "917a3c6c12f967dcd1848b24055095a16e5ad604"} +{"id": "train_06849", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14aae19b17f3bfcf4abb7c6002f2888f3d1bc3ac"} +{"id": "train_06850", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c4a2580477b0deba99031eca3a243c8a2b3900e"} +{"id": "train_06851", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ecda87ed2bb168dc4a494406239b6bc7477d48a"} +{"id": "train_06852", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4add19cc7345adb5edf6479175af44ce7ee71def"} +{"id": "train_06853", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ed7fadb6e2ef2e72b00db96ff46729ac9c6cf3a"} +{"id": "train_06854", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25d0216257188cd008ede9994e27b39a148f566f"} +{"id": "train_06855", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb506d9fed3cfda6180e6439700552033a9e29b6"} +{"id": "train_06856", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5786aa924dd6990ed04d65ca66b5520511b64dfe"} +{"id": "train_06857", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "tooling", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "729ffe325d594f0e62abd7a5aa47737ea800aec3"} +{"id": "train_06858", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tooling", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a22a9a7eb4bcdfe8f5de5b8f9e435cf151975ad0"} +{"id": "train_06859", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50d83363fb252a9b0f39fb7ecb04b01b61529efa"} +{"id": "train_06860", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b49208bc61a8c5f7085ab5118fcbadce4c085be7"} +{"id": "train_06861", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tooling", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8bf0c7e33e75f50f5f53ca58357d1e7141cf898"} +{"id": "train_06862", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "documentation", "auditability", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74a929bbcc53b8ffba75e05c074ec4d56f65bc10"} +{"id": "train_06863", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1111f5e2abdcad816d06c7da042d0b855ae01cc1"} +{"id": "train_06864", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "documentation", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d60e7b9aedff94b2a7cf5afe3bb58e2f5c6ae9b2"} +{"id": "train_06865", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "auditability", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f28d9679ccdcc9902ee4de64760c0a792dc1a02f"} +{"id": "train_06866", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94b13a16538b57baf55c0634176a9013dc5edbfe"} +{"id": "train_06867", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "repo_scale_reasoning", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e693d19ff6b1bf9e3f09c810ab2a1c55c4c86bf"} +{"id": "train_06868", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "auditability", "tooling", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "abb3da8d765df5f6b947c70baef2d7bdae47f205"} +{"id": "train_06869", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d34ad907b4e3a460d29ad26e1b72d1c0e3ed7b79"} +{"id": "train_06870", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31ff956b84aadafdd7caaa8b50126910f9b55031"} +{"id": "train_06871", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdd371b34768ec58ccd17bb0b09f123972ec0fb0"} +{"id": "train_06872", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "494a02624d179f9c3da3852a5817131860892f44"} +{"id": "train_06873", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df84a920525517e34858a39906ee28f8004216b5"} +{"id": "train_06874", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "ci_integration", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7f3fe1a1785fe935856c2b7e8483c8a727df1be"} +{"id": "train_06875", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "tests_are_truth", "tooling", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "704fb6215c894c356c1abe7f0e41fa54b7d1f336"} +{"id": "train_06876", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9450d2673993e9a2bad8ea87538ecb7feb041cc5"} +{"id": "train_06877", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2a76788effe340e2217b7e443fc236d88857fc4"} +{"id": "train_06878", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93ce79aa2e12257f4f442a7338cae3d50644e309"} +{"id": "train_06879", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd7c6b8e7f811fc8b739efd3dacc93fd5f2a66cc"} +{"id": "train_06880", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cb2bde07e51651362a0d8358f68e1d505292deb"} +{"id": "train_06881", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a6b83fac7e9ad52a4d97a6f53f082ec7301c9c8"} +{"id": "train_06882", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "230073e1ba3353e8b2a8eeec3a0a26be6bb8bba3"} +{"id": "train_06883", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34c25e5e463d9ba480c71c48a561179c8a13df8c"} +{"id": "train_06884", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e18904e3eb400b0732a00256ea923b3e88a030c7"} +{"id": "train_06885", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "reproducibility", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c96a5b682e237e36eae507d04ecc855468cc63e"} +{"id": "train_06886", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "reproducibility", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a217d3ce7a745b3533dbd96177d19595f86334b"} +{"id": "train_06887", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffcb273c07aa4ad72844e2568031445c8841f8f2"} +{"id": "train_06888", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d611e720bdaff08a927f2ebaf22cbefd79f57674"} +{"id": "train_06889", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "288f29b1bd60e76339565134f7bd1c98f03ed2b2"} +{"id": "train_06890", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f856e4bc5a75a2c32a7dd000ef986a08e6ddacaa"} +{"id": "train_06891", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "documentation", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a30c68670573a427d378f931b6077612d468ddc"} +{"id": "train_06892", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50db3145f7bed8175a3004d859cae6a9e74a9902"} +{"id": "train_06893", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "documentation", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eae9cd0d1801f5867fda163511603f1287338d46"} +{"id": "train_06894", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13512fa131180ac60930360cdb4a1e7cf4483b42"} +{"id": "train_06895", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "reproducibility", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cd87228170de57c5f25893b4f7cafedebcd6514"} +{"id": "train_06896", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "documentation", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41da74e3dac5f6cd1d7d59afb72b8ed8a95d7016"} +{"id": "train_06897", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0413a264900a05a879c131a72f642c359451d5f"} +{"id": "train_06898", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9ed7923b0b85011101e00a376d029e220f19d2d"} +{"id": "train_06899", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "auditability", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ec5d26aadb0539b2ba3f3304158315418d2ccf4"} +{"id": "train_06900", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b54cc74b9a408a898adb3b78cd5e3987112e8880"} +{"id": "train_06901", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "702dbf27f2f2325b62e727804893f526222944ee"} +{"id": "train_06902", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f195ccc3929806a2e63b96071548b4b8ba169523"} +{"id": "train_06903", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tooling", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8409caf286bd3717b5b597b86949272225a92fa"} +{"id": "train_06904", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8939bf55245cbe6e81230d2f9d70a656fa4c4481"} +{"id": "train_06905", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e94f91a97887745bddb6e5a2221e56ca1637b14e"} +{"id": "train_06906", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d451035841d6421f76fd7bd79d3091f56353b6a"} +{"id": "train_06907", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "repo_scale_reasoning", "tests_are_truth", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4797eba72ffdd1dc33939fdcb3172827708814db"} +{"id": "train_06908", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4414a0bd6017dd9d077f292870bae3a1441fb721"} +{"id": "train_06909", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "013aa2d425fe150eaee87392003867553f97356e"} +{"id": "train_06910", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fad3f353da4a91306839745f47bec11de1944f94"} +{"id": "train_06911", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6f205e6d024a882de1690c80794c8bb5cdec73e"} +{"id": "train_06912", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9a62765f1e7aea704ce24261f8d9a07f26679ff"} +{"id": "train_06913", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "documentation", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6762e79c9a68ea273180876043559ce258666ae3"} +{"id": "train_06914", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6945b99adaeefaff7a0ee96224c7f38853bae2ba"} +{"id": "train_06915", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fd8a3a3c11c6cb07ebfe8d83a2ad1ac18214e33"} +{"id": "train_06916", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc8ec3596a98d295f6f54f89d4e5cf1ab21fdd7c"} +{"id": "train_06917", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "documentation", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9459dc9d18d6b5ffc973ec6b453c7f6de16b62e7"} +{"id": "train_06918", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5c91d672cb6e7f077d012d1a753b8a8b82ada5d"} +{"id": "train_06919", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cba0e810a6ca0f814139614fb459d270eacf69ea"} +{"id": "train_06920", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8f816baee979bac83953ecd8d90f1f18c500a9b"} +{"id": "train_06921", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "ci_integration", "tooling", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61a7ead3504ffe27225c374f9246ec4050cb6826"} +{"id": "train_06922", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "945229df81d9803976ccd7ee88a1b0cd865d38f9"} +{"id": "train_06923", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "224fe3414a2140523a7347870285719259ef4324"} +{"id": "train_06924", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "reproducibility", "documentation", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9083fda08ad7d144521ce9123119dfe0afd72879"} +{"id": "train_06925", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c54e9df1b09871fcc2c0239bd2f340e0d48e6903"} +{"id": "train_06926", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be0f8c5c001f27ceb9a314f44b7ce8c4b03b9817"} +{"id": "train_06927", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tooling", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "075463ebe7e3768c7bc861795b3adbef6227a52f"} +{"id": "train_06928", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tests_are_truth", "tooling", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f446f2b6dbda273e139dbca0b903802e43227702"} +{"id": "train_06929", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "084265739af59f6b24f5d90517aa5f222df2fbff"} +{"id": "train_06930", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6095bad35d40cee1bc1cd57da55153f5004e6ef9"} +{"id": "train_06931", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6311023c10bb7732935529ea66f292c479b8bf46"} +{"id": "train_06932", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25c73a95860dbfe7a346530e4311cb1b556f8f0e"} +{"id": "train_06933", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "security_gates", "auditability", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24c9afba8f17a374ad9364758189b6fd7ec7ae9e"} +{"id": "train_06934", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9627bf1dd51742e3b060c737279c1c7451cb9ef6"} +{"id": "train_06935", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c4956c167e0da185f7ba03cd56054f33a41e318"} +{"id": "train_06936", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7731fcd4401eef47cf8b9b637273857ad705b918"} +{"id": "train_06937", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "governance", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cda31a7e6b142bf286fc9840472132f69931f7ac"} +{"id": "train_06938", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tooling", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b94245fe6a128a53c89cd4c4a914e0013096f343"} +{"id": "train_06939", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "governance", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33b667b9666bc8d2263b173c20d3d8c71d748ad8"} +{"id": "train_06940", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "governance", "auditability", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f42b921c5c9d6eed4dc0a89ff3482228ebf5a0e4"} +{"id": "train_06941", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdeefee183582262f88206a4e347b27d69f2e5d8"} +{"id": "train_06942", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daa46feb0024e4e8e04c20a5b5ff2d7727c3212e"} +{"id": "train_06943", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9c4d3862ce07b80abe53e25855ab27a34effd32"} +{"id": "train_06944", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6346e8bcc641d707539224baf1b8f043d47d21c3"} +{"id": "train_06945", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "reproducibility", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6de0fcce71a404695696530b295adc17db0784e2"} +{"id": "train_06946", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6635da64d705e88321c1feb952cc5a8c24151dc8"} +{"id": "train_06947", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "ci_integration", "tooling", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3613720c041a2a20ea97da6887c030ee7f5716ce"} +{"id": "train_06948", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc2caf6ee9a23d0b37dab141d8d473614432756a"} +{"id": "train_06949", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42294db117eb5ba876939dd3e8994b3275ee80d9"} +{"id": "train_06950", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a04deb969ee213e6efca54bf470dadd7b514abf"} +{"id": "train_06951", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d82b0cb525e5923f16ec323bd84830a515e6df9"} +{"id": "train_06952", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "documentation", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2a9678e81ab5a565feb92a1fe84152363d315cc"} +{"id": "train_06953", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c36bf0b2bae83f3b41b0cb5d1f7e15436e180964"} +{"id": "train_06954", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f42470cf04a6f466119514a7074d77e305e83ef7"} +{"id": "train_06955", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39a7f8e5afca54ec71608ac5f18fce8df5b950aa"} +{"id": "train_06956", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddbd629fa57a6374018fabb8a98e0010b519933e"} +{"id": "train_06957", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "evaluation_metrics", "tooling", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e29d592cd98b5a780fc1209206fa9de2053ac069"} +{"id": "train_06958", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tooling", "documentation", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7877132ce74553ddc17d36442e6e428cd537737d"} +{"id": "train_06959", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b73f5380ce5b25a3545437472520d8e78ab3941"} +{"id": "train_06960", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7d23a629e316459b4fb20f2640ea3bdf579c7be"} +{"id": "train_06961", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d137a42824d07565c3ea287fed7a43b448dd044"} +{"id": "train_06962", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25d0bb9d52c5a06ff82a23f118fb4a1f0b56c230"} +{"id": "train_06963", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd20b3ed1313c2b39416620c629aee7e7e08dcc4"} +{"id": "train_06964", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9fc24e59f14b9f209c8adc77933fb8a945e2773"} +{"id": "train_06965", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tooling", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "530aeed89390e3a3315913911b6447d80050ef73"} +{"id": "train_06966", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7893cda23149f09da239786a9e61a02b71cb6c23"} +{"id": "train_06967", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d89c78acc59f26a0559c5d7ac291d32e9af9667d"} +{"id": "train_06968", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "governance", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcfb392fc468a370d25d640413619433d01122b5"} +{"id": "train_06969", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "auditability", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d15a432003d8980cc8751245d6e5aff9cb76572"} +{"id": "train_06970", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "security_gates", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76af5f70f701944b18b0c00c539b2c450f02fd5a"} +{"id": "train_06971", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5798c3bef9cc2c131327ae6e1e870397aa439d58"} +{"id": "train_06972", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "tooling", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2a20e0af04540166589c112a0b4dd7946af48b3"} +{"id": "train_06973", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "documentation", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b1372c285abfa9d110869ff2601a785a6b93639"} +{"id": "train_06974", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45d562993d38ba4c66c6172060c6ce9c3d520f85"} +{"id": "train_06975", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "945dad6d751245c58125ab03f153b74918495f8b"} +{"id": "train_06976", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "714d523609fb6cd0e039f2334d3dd62619fc2a85"} +{"id": "train_06977", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "governance", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "469a14d3a59d0f5d5202893c7a299b87aad54df7"} +{"id": "train_06978", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "documentation", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "585670101266adb4dfa3360a0f001299935fc190"} +{"id": "train_06979", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "governance", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27b20287809be0d9d5f5074a855d02ef91b91215"} +{"id": "train_06980", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "704fb6215c894c356c1abe7f0e41fa54b7d1f336"} +{"id": "train_06981", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a868d54d54a242f971991b4173205bc50245a696"} +{"id": "train_06982", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "security_gates", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "399dfa3a2a71292c6ce9ba7be3954762ccd46563"} +{"id": "train_06983", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tooling", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9199706d032336d99a83682ce320351629a7eeb"} +{"id": "train_06984", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "governance", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b030f28558c2952e41394c69e1a3a2939ba1adfd"} +{"id": "train_06985", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tooling", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43a5989499f76bd9f1c62651e0ff6d4a932362c0"} +{"id": "train_06986", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8514a0678be3cd517266949e266b34850f42c1b"} +{"id": "train_06987", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "382e81da81410414b530437bd99e4fc30a5ad278"} +{"id": "train_06988", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f86a160aa999d9aa5554e0cddce18f1bab9c902"} +{"id": "train_06989", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cf65ef710c039054f5f12a2b2ca968b73d67845"} +{"id": "train_06990", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "ci_integration", "reproducibility", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ddf9e298df2bfa7beb25f956916504fe5e38256"} +{"id": "train_06991", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7e60a2a5a0a0dda42fccbd4b3c32f5819bdbc2a"} +{"id": "train_06992", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9663ce9ffaf44e7ff6c51d8aee7bc77ecf64d7ca"} +{"id": "train_06993", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "evaluation_metrics", "documentation", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2786b44628ea012288cca682556582f43d82cd3"} +{"id": "train_06994", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30d5b412b52e4933b1f2158e5b75479d7e4b729b"} +{"id": "train_06995", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "321d4cfcd9e0918650f77b9be4790a2cc744c493"} +{"id": "train_06996", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3a7be1615f13557524072c9df6ac0eac07b9b12"} +{"id": "train_06997", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "governance", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dba8452eafbb660ddf0bf35eb15f5bc593401054"} +{"id": "train_06998", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bae99aa8afb085fa0b6fca61190df915195a992"} +{"id": "train_06999", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8556c2808dbed540f43f545ba3751cc772d18c1"} +{"id": "train_07000", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d289f915033d4e38e694376cd1161ef0ef16034d"} +{"id": "train_07001", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07cad43701b999da8538a7506ada6893088df76c"} +{"id": "train_07002", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdfb22ea425157427fab54c53a9625a39240975e"} +{"id": "train_07003", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f82d37063ca2f40b07479563ff1ddd0725fad8b"} +{"id": "train_07004", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9478758d2f704cdd46224563f4d1b821aac87de0"} +{"id": "train_07005", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tests_are_truth", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d57c8c65901ec53d8e0ed53ffa764ebd32cc8513"} +{"id": "train_07006", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ed349a3a377b42a086a30bfc8f04a1a9113762b"} +{"id": "train_07007", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc871a73baab89eac9a468d87f671832351f1b6a"} +{"id": "train_07008", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "reproducibility", "governance", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5543690a5ffc95b3af801467ec7a98a8e7c9e576"} +{"id": "train_07009", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9aa0a7a75707a078534560c5cb3a28646163d2f"} +{"id": "train_07010", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "reproducibility", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "395ebeecfe5552bb8c4d0c03fb6d62a4175be94b"} +{"id": "train_07011", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8b2a1fd9eca68baec40ec31ee4b176573e66c74"} +{"id": "train_07012", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ec4fb6ffbd5fc303f8996bff3ec87093ea66a4f"} +{"id": "train_07013", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "reproducibility", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "907453efedba7a789a7c3b711fdd39e09803883c"} +{"id": "train_07014", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db6c1d69f14cac1b73f448c433978488c5bc2f8b"} +{"id": "train_07015", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d57b5392c1611bc37b5673825f1c37ab4e54922"} +{"id": "train_07016", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "552c541068fb117f30dd2805f9c92344cde22bb8"} +{"id": "train_07017", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "098d4a6de045937b186627fcdb910e6bdfb0538b"} +{"id": "train_07018", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "ci_integration", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1369752cb8b5b108a50f80e90c8b85dfb791e8ec"} +{"id": "train_07019", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d8e4349fb560d2cc9a89b09896820565109c138"} +{"id": "train_07020", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d37c7c8ae68b5905da259d53f6a8e9464db2d6e7"} +{"id": "train_07021", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94d82f3228f02146a38a974767b2bc57108361ab"} +{"id": "train_07022", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "559ccda68ba4586d79396f350cfd81a350f7200a"} +{"id": "train_07023", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "documentation", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88b785d3ea42f4f2098cad00201321b5092f43e7"} +{"id": "train_07024", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7594fa8fcd18fbef21c4e6da5e9e9822a98c79e6"} +{"id": "train_07025", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35ca6a4eb258d972894e5e0bbbc01760db157d42"} +{"id": "train_07026", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2971397baac89d3d00cbc258453f08cee33f7afd"} +{"id": "train_07027", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "tooling", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b993f714a59d146fd79940306b6dec267a22ecfe"} +{"id": "train_07028", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d198ab8c1c9fe6199a855de8bebce27ee402bf7"} +{"id": "train_07029", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "tests_are_truth", "documentation", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a189c4af678cb3dd3803e313ac87357c8bdc790d"} +{"id": "train_07030", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c70a6f44075c57b3df829065a346f7de60872c24"} +{"id": "train_07031", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "ci_integration", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9f92388fa1ca80e8b782756997704af1a92959b"} +{"id": "train_07032", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c072e4dab116010a007e1e167f62f3685d0649e5"} +{"id": "train_07033", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2b7dd443b9fea4ec0e28fdb6c91b5449dc72206"} +{"id": "train_07034", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "repo_scale_reasoning", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a80dd3e3b3ea37e7a7139e55f06b67ea2c0adc9c"} +{"id": "train_07035", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0125a48d7a92afb403a7d162ea2df7df73c29f9c"} +{"id": "train_07036", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "reproducibility", "auditability", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cdc46bc3cbdad20901ea8cd0d9f11a0cc8f8489"} +{"id": "train_07037", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40bd5bad497566aefc6434f6da2aea808bd85d92"} +{"id": "train_07038", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "auditability", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d4583d24f5aa5c7680d4110d387da9eabf02a5d"} +{"id": "train_07039", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40b3ebd9085be3950eb92461cdce15ec9531c05e"} +{"id": "train_07040", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "737ff1ad97255a9cecf2fe17ffe1b77c9ea5568f"} +{"id": "train_07041", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99ba2b335cff621c86a561ff3b0406fd2f6d8f47"} +{"id": "train_07042", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ab18a19999d208c7340030022efa2651c850b31"} +{"id": "train_07043", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa1b52e95bc5703cb25c8e882055fa466fbb16b3"} +{"id": "train_07044", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8468fb4c4e7f31afc0a1a53b9ee0be57c1676bd8"} +{"id": "train_07045", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b7404dce9dd5b1e5a21da7e61a8612aa4c074df"} +{"id": "train_07046", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "528c3d76c0d4bfad514a7ba5dd656cfc8cc97120"} +{"id": "train_07047", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "security_gates", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8aa72b8db1d2abca8e1db8e2203bfc0801f15ebf"} +{"id": "train_07048", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78f3acf610a47a4870d5acd5cd80bb27673ce452"} +{"id": "train_07049", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cf659b223a8c271ac77e90cebdb99d43a43c32d"} +{"id": "train_07050", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfd6ca8197c22169db04d67af567abe7d2169f40"} +{"id": "train_07051", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a68c721979b69cac104f842a0ac93d27a5e4a8a"} +{"id": "train_07052", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5d1c9ad47508f4ec0a87275362786e2d5ae28f6"} +{"id": "train_07053", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef2f1e6a96b361c76b5a06a8d10d331b0a9c5f0"} +{"id": "train_07054", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "ci_integration", "auditability", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31be1fd38ab39e1ff2f5a41f6039abf131f78d6c"} +{"id": "train_07055", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "558c308f0ce2dc4e77515daefc75bab2cc3c6cf0"} +{"id": "train_07056", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7105e3cdf06d307d66da4a099449dae367ca0478"} +{"id": "train_07057", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e0983256e69717b5018610d52eaf4b71f2f960d"} +{"id": "train_07058", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a9e156bc5c95c12ea263b4c50c91744e551e2e6"} +{"id": "train_07059", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "documentation", "security_gates", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b0c32c134fb602bd452ee79258c1aa3b5d2b2d4"} +{"id": "train_07060", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "auditability", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af0a1ec68f5fb69c31aa0dd02f01824b6a810a07"} +{"id": "train_07061", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59710642b29fe7278e0a6491174a517ea82c025f"} +{"id": "train_07062", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be22701009bca377bf00fb01a84b56f8d02a88f0"} +{"id": "train_07063", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfc215b7bffe58408ea1dd9af62c1799461364ec"} +{"id": "train_07064", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "auditability", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65c9aaadb6c1e73cf4dee1cd486de6122f0a4457"} +{"id": "train_07065", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "tests_are_truth", "reproducibility", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1657d60aac269de4739ac3d1ba95de1b2bf94c0d"} +{"id": "train_07066", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0441179e0b10f8d30cfced57689c434b6941b078"} +{"id": "train_07067", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "governance", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "abb25833c194b5e7a3ba846f654d8ae6579e5f1c"} +{"id": "train_07068", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tooling", "reproducibility", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16d7cc0aab968b3845ac692a7c25177462e86afd"} +{"id": "train_07069", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d6d5c19c7b0f0372de8a8d016b4b75a311dd21b"} +{"id": "train_07070", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d65e0764e75db68fb44bdf6138e216466d320213"} +{"id": "train_07071", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "repo_scale_reasoning", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7eec7c8baa3362c38dff58def4651dd1d5cc108a"} +{"id": "train_07072", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d67de2d229fccfcde35a90cbd2a2c2d0ebabf0d"} +{"id": "train_07073", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5fde4917b7c9c737684f8fff02ef9ed17e3ec05c"} +{"id": "train_07074", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tests_are_truth", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5097b112a46e957a605a5224d1d59fb17ddf0ccb"} +{"id": "train_07075", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "auditability", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c49a66533eebf43e07c941c80f60a4d788eb2283"} +{"id": "train_07076", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53273da4a7b20f8d5bc1f36025534d0175800f3a"} +{"id": "train_07077", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09d90524b6f6ce82e48abebe5550f5dbf5ef2123"} +{"id": "train_07078", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "documentation", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79df8183b465d9f5cf37ec3c076e8ca68e926689"} +{"id": "train_07079", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "reproducibility", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1369752cb8b5b108a50f80e90c8b85dfb791e8ec"} +{"id": "train_07080", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a6b8914267fe52193edbfdfb78034884c2d46ab"} +{"id": "train_07081", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "governance", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e691464c2582655d26f1db4e376cbf8722ad7b6b"} +{"id": "train_07082", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f4a6e15f551e10fc82c33d5ba8a2faf938dd15f"} +{"id": "train_07083", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc23fda4c2757e055b488fa7010c63fd2b72189c"} +{"id": "train_07084", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c15d62484936b9da7a65aef5af872e7108926d57"} +{"id": "train_07085", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85695962e94b6b5b95166a44a13910c30a56ffce"} +{"id": "train_07086", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c541cfad2747a1b52c2a449e5a24bc8153dc4eb0"} +{"id": "train_07087", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "654e23ec3772c9f2459273b4adc9ca70ebc0220f"} +{"id": "train_07088", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68c1fe96fd1a6b46a6253607446ec41c1da7356c"} +{"id": "train_07089", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0101aea55700892f232170a96b53e48868b1741c"} +{"id": "train_07090", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2d9d97ba66139ac53d7040343408a0e9992efcc"} +{"id": "train_07091", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c4760783ecb46e71ff3786e8b60d98c499687dc"} +{"id": "train_07092", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d6083856cdcabfe1af6e5c4914ff25cd28d62e2"} +{"id": "train_07093", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "tooling", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "411fad2efda8631649dd40235eff42f7ef66a309"} +{"id": "train_07094", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "auditability", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db38baf1c63c8e94f7199210d5ac3d23500073fe"} +{"id": "train_07095", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c982d3fc7f1f618ca19c7b0d2e726e58ffa5d59"} +{"id": "train_07096", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e5e1353ad147f949bd5f0db92046829aecbb99b"} +{"id": "train_07097", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tooling", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d87e9024f4e7fa95e57469b7f4ae8f1c450e51f6"} +{"id": "train_07098", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tooling", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12736403ff5eda6dd0c9bd50ce46d2c9711cd2a3"} +{"id": "train_07099", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a594d798003d28e1a271d719f1cb433f9ff02c6a"} +{"id": "train_07100", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfba0725dbef3f39c838c055a214a1d0e2a054bd"} +{"id": "train_07101", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66ed419da1c55e0ad18cc9a61e68626991619c37"} +{"id": "train_07102", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df0a95a24222c2c41295c413f5712b5f4531f120"} +{"id": "train_07103", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8982f5c1093ead68741db1c1b01c22c8d1c22d42"} +{"id": "train_07104", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "ci_integration", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24dbfd66ca72b915cefeb62ea0b6fbf2bc90d60f"} +{"id": "train_07105", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62936cc5ce4ad14fd71627f11c8a4ee736f6d830"} +{"id": "train_07106", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d87844f09f7425293ecbf4b3d7332df049c4d1a8"} +{"id": "train_07107", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "documentation", "auditability", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11dba07b6f06974261a27f29f807f617cb1c2ecf"} +{"id": "train_07108", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "ci_integration", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0923080ed5073f3f5e2692a4c08cdda61585390b"} +{"id": "train_07109", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "governance", "ci_integration", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9f75d5829264fe3f69eedd2f92dedb182531074"} +{"id": "train_07110", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89429d6948ed68f726df74b29afce98471d1dd5b"} +{"id": "train_07111", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52c3d9492fe3a2405fb50e45271ee98116fc0ef3"} +{"id": "train_07112", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tests_are_truth", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4990b4fc17d5f182b30d3ae8b308f9d54304be1d"} +{"id": "train_07113", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d90244907183d876c52f3b30ca7e5cdf628d3dff"} +{"id": "train_07114", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86e8feb58091ad990a926affed7fa3e9e6d3a5ff"} +{"id": "train_07115", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ac2b4a05447038b3e4b2095112735ba8d14c986"} +{"id": "train_07116", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83d657d624e28c2c7af1f46c1213594434e2fdba"} +{"id": "train_07117", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb4163661b4ff4a8ae307f6e7179dab34305a33b"} +{"id": "train_07118", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tooling", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1792a90830716f1873cf0297874af708ce6d93de"} +{"id": "train_07119", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5372814ac33df70157b3abc1058e73b45ffae256"} +{"id": "train_07120", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "577b342fd3de46312ec73f73934b93f2143b4398"} +{"id": "train_07121", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a970e16c5d4cbe7655a419b8c5968bc1b15e921"} +{"id": "train_07122", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4b58af8891e9f860120f66738ecbdeb3b3b30cd"} +{"id": "train_07123", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "auditability", "tests_are_truth", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f7b2f8730c8ce252fc697ff5f3aea7e498a08ea"} +{"id": "train_07124", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a418b363c637ed3e16c7678f2dad36ed537d74a9"} +{"id": "train_07125", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc512f9d77d213bcc84f8854dc3d4082371054ae"} +{"id": "train_07126", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba35ea621a2d76c72cbd5936dd46daac1f469b2a"} +{"id": "train_07127", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9326e1d5b7be4d0cfb3eb36f2aab7d0273dbec22"} +{"id": "train_07128", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab27f7330d6edf4bed89284351682a81079cf3d7"} +{"id": "train_07129", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tooling", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9faf4e3d7876d89b5a829cf5ba31e14e6d88a46b"} +{"id": "train_07130", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc42d6ccda679bccdfe3c7f03869dc85da2f55a9"} +{"id": "train_07131", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "ci_integration", "security_gates", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77dd33830800edf6138b0fee69b5592c3d03701a"} +{"id": "train_07132", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "tooling", "documentation", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c1fb858fcd83621123ecaf5098a908901eda622"} +{"id": "train_07133", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "governance", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c77817b5ae8d9ef200dbe2f49045a4422f354697"} +{"id": "train_07134", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "governance", "security_gates", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd394ce393bf5f59f0f2906b8d66ca2811b80937"} +{"id": "train_07135", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4de82b71fc04630712bc98bbe131eceedcd378a0"} +{"id": "train_07136", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tooling", "governance", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b59d255fee841b72cebb5b43c7493c2bf68aa6d7"} +{"id": "train_07137", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "tooling", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3643b3e7111aa71199314769f759753e6c956dbd"} +{"id": "train_07138", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "governance", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5c91d672cb6e7f077d012d1a753b8a8b82ada5d"} +{"id": "train_07139", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6eb5e336e4a75724fcdc33e27cbd6f1904423045"} +{"id": "train_07140", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "documentation", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5860597aa1dd545e1885ed010c2e5f05cfb3944"} +{"id": "train_07141", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96612f77aea42fb7a6e5709cbfc775349d9355b9"} +{"id": "train_07142", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5391df199813e14c5cda6010bf51ecff677e2dfb"} +{"id": "train_07143", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e4509c248e774e0ef5e0c3c6b88383dbf96eb3a"} +{"id": "train_07144", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f790ce24fdd29fa1b32f35027c700d0bcd89064e"} +{"id": "train_07145", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c80147d88cd04674969dc7bb8fe53b5f262a36c1"} +{"id": "train_07146", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93748e94f71390610174fadb834e484f0bc99e62"} +{"id": "train_07147", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1717eed285faa02e565d7a0e07578351834ac33f"} +{"id": "train_07148", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ec06510984c4ffea71fc6af10bf62ff341c0367"} +{"id": "train_07149", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccb9b52ff5786742793388ddff563c44d92e8b2a"} +{"id": "train_07150", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94a4c8b8ad307bc4f0ca31e1004284011444dfc3"} +{"id": "train_07151", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73cd431338c89dcffd130aab74d9f583f5de6f37"} +{"id": "train_07152", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd1411b5786e21d6d105c366a70a43641303bcc0"} +{"id": "train_07153", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "security_gates", "tooling", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34ea5ab3789941bdbd5a16636018459acb529f1c"} +{"id": "train_07154", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bba7909337adbf4801578f5cf865a4a9d610b32e"} +{"id": "train_07155", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b97edd173f1589b226cbaf1e231c364947739ad3"} +{"id": "train_07156", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "reproducibility", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cc92dedf1f08512af6d4d27f027921f071a52f2"} +{"id": "train_07157", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ec3b3e991a5e1aefe87c1d186afe882c5d3bf99"} +{"id": "train_07158", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "auditability", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b539e6f38883ab91d86f725984ab5e83aaf4dd7d"} +{"id": "train_07159", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "documentation", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72aa1c499c543c9e51f533f97d2d93a3677c5946"} +{"id": "train_07160", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "reproducibility", "documentation", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5c347dedaf7a3a1030b58107f4da2a4f8a19360"} +{"id": "train_07161", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5712e5a6abda5fb9e4c877b0d6a723b8be7d1cee"} +{"id": "train_07162", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "documentation", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a26840df9a2b8e6bf51603f7289b4779394bf0d1"} +{"id": "train_07163", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "reproducibility", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fb16dd2ae8fc8333ae10570362e89b0c7962122"} +{"id": "train_07164", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69c40552dd2cb1290d7a3bc091aed3c3c90f77a2"} +{"id": "train_07165", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a624f34d119a4ce9c289f708aeac91250e7c36c0"} +{"id": "train_07166", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de549bb6be831f8bce57d7e474424a27692eed69"} +{"id": "train_07167", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd9e222d23e39392263e3d3f70bb98bf5563f1db"} +{"id": "train_07168", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff7d673bd9148c6ddc58fdfd3ad142779a946ce5"} +{"id": "train_07169", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a9be9cc0999548d743aa97b982d2a766f7299cf"} +{"id": "train_07170", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "governance", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "535fa8df1b682ecd3c3b6601b586a691459fb9a9"} +{"id": "train_07171", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "736b9eacfbd5a515b8db4d260d164b37623cd266"} +{"id": "train_07172", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "auditability", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37cdeab78224b7c9054c7dbcd742d2dd06a3090e"} +{"id": "train_07173", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "security_gates", "reproducibility", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc21f38243be8b6430c8b7159a8483989294ee09"} +{"id": "train_07174", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c88172c0836998a3106a750fcf0c4737208fa8d"} +{"id": "train_07175", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "ci_integration", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f65fadb83029145f7cb617ff0ae79e1b247225c3"} +{"id": "train_07176", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "documentation", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8aebc4b2ec46147ef1a3ed5c78ec97850e8cedaa"} +{"id": "train_07177", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d745f842aaff0cb53bb09e8335cd46e025ae338"} +{"id": "train_07178", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "auditability", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ef04aefddbe8f3db835ffe50dc8634ad9f34fcd"} +{"id": "train_07179", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ad5926bd04ab4d880cca9802ee28a7e19d0157b"} +{"id": "train_07180", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "374b45f269446b6ccf91d9e1731c522767d8cef5"} +{"id": "train_07181", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ca141abe2c1a8641f7de1796c417424944c2a9e"} +{"id": "train_07182", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bdf5f454e32fb6f10b4b58e437135d728821527"} +{"id": "train_07183", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5014eb56809eb114706ce5443909f6101d7ad443"} +{"id": "train_07184", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "886675c682ecfe80ed597f704f9e80411571f1b4"} +{"id": "train_07185", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ea246326f559030a5ac78deaaae3ba0174ba559"} +{"id": "train_07186", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3555953b923d75143f425a4335a2a64068b0cb5"} +{"id": "train_07187", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e76dd0632106e71489f6baa71aadd813059fefc"} +{"id": "train_07188", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "governance", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9450d2673993e9a2bad8ea87538ecb7feb041cc5"} +{"id": "train_07189", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1592450239f07cc8e4be0b22e3dfe9fabe374522"} +{"id": "train_07190", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d31b606693ae7a0d12af5f473016587b169bf86f"} +{"id": "train_07191", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "documentation", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9ba0aa6425babcaefcc0ba030f0b5901b492653"} +{"id": "train_07192", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1a6d43be1e78ceff0d95bb4f92dec2eb92138a8"} +{"id": "train_07193", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "238640525c7ffa35379d102c2e177225b5809756"} +{"id": "train_07194", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bf5420631afca8bc8af271165b6b1fb0cfb6c52"} +{"id": "train_07195", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5fe197ed3aed4fd3b217debe5775fcd205e120db"} +{"id": "train_07196", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7abc592e759cdbabe7849351af296020556de1c"} +{"id": "train_07197", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tooling", "reproducibility", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3c6a52e620a321abf62ec9f5e86e9bf0f3efa40"} +{"id": "train_07198", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "governance", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7c33f9521ef005d58cfb18669e536c1097d7cd4"} +{"id": "train_07199", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "documentation", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7e37f1a020f3539e2759932f9c6507fb4c57ad8"} +{"id": "train_07200", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71884849d0420b4d8e387fda9afd3ccf69ad29c4"} +{"id": "train_07201", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d707a47ff334a25d25c2d69d9759b0bf2a02c9fd"} +{"id": "train_07202", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71f6f17c00fd19ede2bc028cd909f994668db607"} +{"id": "train_07203", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "security_gates", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f4319e8979d51d95793941c280f0eb81b267e0a"} +{"id": "train_07204", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "auditability", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d614ba08a0f36713fd13e7fb3b5820e7a854f80a"} +{"id": "train_07205", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e494702397a5b3200556ad7abfd14d439dcb960"} +{"id": "train_07206", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3424a71bfdd07d575f26e86218a3f477209a94f"} +{"id": "train_07207", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcf557f9d5b1c43008ac94fdcb2c510d890b5310"} +{"id": "train_07208", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23a6549553e3d1fb8c483fc3ac801cb33797f4fa"} +{"id": "train_07209", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "tooling", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e88cc3795c101a46db2b871dcf5866bb3080453"} +{"id": "train_07210", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b699c470304cb69e9566f3f8ed9b54a42a8ebcc5"} +{"id": "train_07211", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed44f3dea80594cdd74bd5e08a5c58051b398c2b"} +{"id": "train_07212", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bae99aa8afb085fa0b6fca61190df915195a992"} +{"id": "train_07213", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8c5b9d0e49023e5e8c19307847b7e2a13814d5b"} +{"id": "train_07214", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff76105755eaa066094c491bc87cbe82a7b987bd"} +{"id": "train_07215", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c57f739f3c0a7c357e43b0afc06b9831f8e2306"} +{"id": "train_07216", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e86a9730e83dc439f322802a31826a1666deb11"} +{"id": "train_07217", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e296fefcaa19d5fe107a2bcfb834063b1364da0f"} +{"id": "train_07218", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec7bbaa2750da6eb783bf4511ac59c2b7a8b3885"} +{"id": "train_07219", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc21f38243be8b6430c8b7159a8483989294ee09"} +{"id": "train_07220", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fa156ca94506031b6b36b176935cc921beaa006"} +{"id": "train_07221", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f90fe55485d1b63536542dbc9906196aa77f6e7"} +{"id": "train_07222", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "governance", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0882456aa52f0aaf9004964a4dc439a3a4fac3d"} +{"id": "train_07223", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0351aea923f9cbe0b262b94f0bba28af043fa3bb"} +{"id": "train_07224", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b2fe69af185048728b40709dea4fec76bef73a8"} +{"id": "train_07225", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ac43057c2e2da09bd5c79a07de653aa480952f3"} +{"id": "train_07226", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5752fbcd3793911cf195d52ac1ef91756fac9149"} +{"id": "train_07227", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c77e79343dbbe4337c29aa40f47bc36b91152b7"} +{"id": "train_07228", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "820d6353605c686c9a0ec8a7d8644445c2442c1f"} +{"id": "train_07229", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "reproducibility", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93bf85f66d3241037d7d17fceaf59e8815932bb1"} +{"id": "train_07230", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49f62a78f766cff7a5ad7fdfcd69650e245a921b"} +{"id": "train_07231", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2296125e523c79f7a9fa5d4b42740b57a40b88f"} +{"id": "train_07232", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b50b84a7e6aaec5dd37b6441151d1fd90734591a"} +{"id": "train_07233", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c45174ac86c32babfbc685098d1ee9f0fddffcd9"} +{"id": "train_07234", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4c690f878579e00636807e535bda6aa4c631adf"} +{"id": "train_07235", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e20c64c00224e99cd0934b5fbb3d0f87cdbbf9ad"} +{"id": "train_07236", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "auditability", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e171b6dd610e0b47499b83c1c85fb70a5126e26b"} +{"id": "train_07237", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "ci_integration", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de79e4e19d8ee9e7d4863add2c40dc3b715c112b"} +{"id": "train_07238", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58ead36d428e5b5364fc461f70c9c4823e2b17f1"} +{"id": "train_07239", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "governance", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f610c852a4140d0237cda85534561b0f7ab8e03"} +{"id": "train_07240", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83aac36901f21378835a5f4e3ebd75df88cd5ac6"} +{"id": "train_07241", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3712a8f19f2479cafb23d9abe86eea321ff9620"} +{"id": "train_07242", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6d02d68514f5367b480b3226daf74c1f5c58ab7"} +{"id": "train_07243", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f06c81d9b7751ee419fd09a88d00c470f404c67"} +{"id": "train_07244", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f30244dd7159d32d78242629de68a434cd4aad0"} +{"id": "train_07245", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "documentation", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e86364c0b25ce7dcf5f49eb523154450a5e2c2d"} +{"id": "train_07246", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "ci_integration", "governance", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31b7d7648d71bccfc499b5fd2498c16486a9d4ad"} +{"id": "train_07247", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tooling", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f692e7be9925b729e092530858ce3f841e948592"} +{"id": "train_07248", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff2ede8739ca7cc0059cc65860d8f25d52791d9c"} +{"id": "train_07249", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "governance", "security_gates", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2f85a69bee67c048405e37c6475f1fd07da85fc"} +{"id": "train_07250", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47a3095756bdeb6a011802700ca47d4258d78ff0"} +{"id": "train_07251", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9768452fbb74886295779a4a8f0ba320fd2aeea"} +{"id": "train_07252", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72b9500944b4ef5e5e70f895dd266789fe4cbb4f"} +{"id": "train_07253", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b549417db1ae5bd501c8f07456422f7a2841d5de"} +{"id": "train_07254", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "evaluation_metrics", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2fcad05413a410076bae7e518b15feb9f86e759"} +{"id": "train_07255", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b159156b86bf87c4b2aa608faff125ab0410cce"} +{"id": "train_07256", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "ci_integration", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ad8499d8a36d08b32bde056c894f38b6851e8a8"} +{"id": "train_07257", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06fbbe411dc38b93aa158251221c60802689c7c6"} +{"id": "train_07258", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cd17f4452817027deb83d21b86fe3694ae91044"} +{"id": "train_07259", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ff282b72231711815e0d3eade06df465f1ec9c7"} +{"id": "train_07260", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85472f62bb83733a0b98b8bed7cee281bb43dbeb"} +{"id": "train_07261", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d4cdb322d76eb3bf8d65cceda4b87c6ebdf9b5a"} +{"id": "train_07262", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "562289edea085b49f241c78b07f3f75016739dfc"} +{"id": "train_07263", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1be6c173af10422aac72eeec9a9ea5fb17333863"} +{"id": "train_07264", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b9996f8a31b75dd272152482f02cf6ffd6d000a"} +{"id": "train_07265", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a41ff640f8c21804bb4aa4603fe7f225e75048c"} +{"id": "train_07266", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9567105427f993d78e18f7ed95a038cb021efa47"} +{"id": "train_07267", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b419d10651029c143eda254019ce44c20c97ff0f"} +{"id": "train_07268", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "auditability", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b91111cbbdd4de8ae201572ac3f99be20a3d118"} +{"id": "train_07269", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "documentation", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88a125d1621b1aadd30eb27ff414028df32cd114"} +{"id": "train_07270", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tooling", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ac06d994ba3099295a34b06ca67d9e4d1d9e1d7"} +{"id": "train_07271", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "tooling", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b56d0f41b686ff142ed422b0316ef56c4a9440c4"} +{"id": "train_07272", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c93ccfc98b1f01e970c3b7303347d5971d8b385"} +{"id": "train_07273", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b2c6717d4d346141265daee7aad125371e636ca"} +{"id": "train_07274", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66e62ac21402a556365901f2c448c9d37df18f12"} +{"id": "train_07275", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8636f453bf2c6a18cac5086e812318118df1a108"} +{"id": "train_07276", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63cbad3cab0401039293db79fc41b99502566483"} +{"id": "train_07277", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tooling", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6642da0a2409148c74c786d7e5b34a2391fb770d"} +{"id": "train_07278", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8a8f907e55b5b935fb47a21bc8b65a6daed9f9d"} +{"id": "train_07279", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c75c543333e41b4a30f7ff4555e4c468dca3bfe"} +{"id": "train_07280", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "ci_integration", "governance", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8fe15aae8c00e7055050a921e9b07acdb59c2520"} +{"id": "train_07281", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e4a7c6a09573e6723935c58ef3fe77f5c84eab4"} +{"id": "train_07282", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91722486214f4e38efe08c988f1757bae46711dc"} +{"id": "train_07283", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "679a23971f4c5aed20525577296965349bd1c4f5"} +{"id": "train_07284", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f9a36b8c8f4ecb8247034921d79c94d936d30df"} +{"id": "train_07285", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a6dbdff6f384dbb9d09ee5d47badadd6dffae00"} +{"id": "train_07286", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a44c376efd1863fc74d0bd8c1a2697460ce25918"} +{"id": "train_07287", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1535c4fe7009fc7cd8429623b3180d4a3bca7033"} +{"id": "train_07288", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "governance", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a291b431b91773f292c421f4c35cf99749b71b3f"} +{"id": "train_07289", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "documentation", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68df4ff8a09d9b3d2866aec98c58b1a51b9b81e3"} +{"id": "train_07290", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6744f99415ffbc33be867e82619536ec3e62bc51"} +{"id": "train_07291", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe2c05e78551f41c2abc81ef4a9c2cb687f53ba0"} +{"id": "train_07292", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac765edb38930627533bc365a641f85cdc113606"} +{"id": "train_07293", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "documentation", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfba0725dbef3f39c838c055a214a1d0e2a054bd"} +{"id": "train_07294", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "ci_integration", "auditability", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c891defad4c09170056b2348d6fbb6b80616f5e"} +{"id": "train_07295", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "documentation", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aabe3cbddfd28160c37bd262e6cc9009382dd16c"} +{"id": "train_07296", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a03a09131207a259825c6b5571702a9d8881ba78"} +{"id": "train_07297", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57741dcf1d88d5fca84d199d803f90635d82ff8f"} +{"id": "train_07298", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70aec5b7194ffb925ced89bc0a278a4169b9c684"} +{"id": "train_07299", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9552732c8e53fa9b06f34b1a311ef88833a0d077"} +{"id": "train_07300", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "780f048f58441978dd64234f59b136f7098595ab"} +{"id": "train_07301", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "461283839b6a1260d611379c35794a9d9533f543"} +{"id": "train_07302", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e21da65bf175ab076a3cf2e57bd39cda945b5b6"} +{"id": "train_07303", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bea65a9e3cb79e59235e7ec81e011e50113c43d"} +{"id": "train_07304", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5e3e204c8d1ca25bcb7cbf8f02d5eec46c70999"} +{"id": "train_07305", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "970a4cbfcf5dd91b2d5e6348b704c5ea0752d0d7"} +{"id": "train_07306", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "632746ef806ea48679127fb3ef0051e32a3091bc"} +{"id": "train_07307", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tests_are_truth", "documentation", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bc98b2c772b586e02cc9fa4a14f166c32b37c1b"} +{"id": "train_07308", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ea46c5aa08a0f2fc11879a300a0764b68f3bbaa"} +{"id": "train_07309", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c635e09824c880587f1556789b2fbe4a7565ba2"} +{"id": "train_07310", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "security_gates", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "453c6647ae67b106a693a1a022911e6b2bd0ad9e"} +{"id": "train_07311", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67d7007c6a4a069dfdc460a9bb9614a0eee59253"} +{"id": "train_07312", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a37660f8516047c9c633da9f137b0fc43d7e061"} +{"id": "train_07313", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b521db8b4050d8368443cdf307332a0cb70a72f9"} +{"id": "train_07314", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "security_gates", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4523bab3d6f137c7681a1b8ff1cb00eddfca8878"} +{"id": "train_07315", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca230a0fc821b9d2f3bc08eddb71ba92926d05e6"} +{"id": "train_07316", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b91c7db2d949750de8cae6966ff51db2d044a467"} +{"id": "train_07317", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccbe01e1ac8bf984f15b7c05348e19a73c4ed51c"} +{"id": "train_07318", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50754d607820113116d94a597d90140c1ebf5357"} +{"id": "train_07319", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "security_gates", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22794a233163b37a3c2e71ee822cff585b6971b1"} +{"id": "train_07320", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "auditability", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "925c5bb6ce92b69909db0ff6b1aeb7157b54053f"} +{"id": "train_07321", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3f1926552eee26ab727481f9efc305a3b5515db"} +{"id": "train_07322", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71e548e3a56763a5ccc6a90c4f8430eed7f255c0"} +{"id": "train_07323", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19eb337c6f85eeb93641e21e5c656f477c70af36"} +{"id": "train_07324", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "ci_integration", "security_gates", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76b9174e991395f2be3ecee62a6b0944848bfce9"} +{"id": "train_07325", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2f8bb7fa6a2516432f8afa285efef1cd6e6c322"} +{"id": "train_07326", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14b4f4c24241b3ef6fc32300a717f8e5f1b850d2"} +{"id": "train_07327", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13d2000947543a8db080fcaf06c022c20fb1950e"} +{"id": "train_07328", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70765aa0c0f677a962b90deb950f0aea694a59d1"} +{"id": "train_07329", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f33d4d7d3c0d178787dc88a3e0646650c76c5199"} +{"id": "train_07330", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "auditability", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db0c0dedf8bf61bdc8f589db18f94d0fa5f204ec"} +{"id": "train_07331", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "governance", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f50791261ec1039acb37781abf6749a13d1dbfcb"} +{"id": "train_07332", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b04a0c1b0645a5d4222072492643561858f106cd"} +{"id": "train_07333", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "governance", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6a20c482766e1fb7c9abe0f85fd1ca0455a8437"} +{"id": "train_07334", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4e25b1aec77004c5d52a83805808c0c9f69d4d2"} +{"id": "train_07335", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "governance", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a156d4bb5ebd41fc1c8627fef39a764adb866a8"} +{"id": "train_07336", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83c35ba9bc239b65352df66df7168599f51c2f2a"} +{"id": "train_07337", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "documentation", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d92763bc72b91073f51052543f40176ef6deb0a2"} +{"id": "train_07338", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "ci_integration", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f443d8ecc5fbc7702feee6f5c5ab9899e279223"} +{"id": "train_07339", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1125592db3f66599ee8dc948478911307c916c01"} +{"id": "train_07340", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "reproducibility", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "526211db4c366083cd4b243809f19c5d7976c722"} +{"id": "train_07341", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7baf01ac0d8970e5913260db7dd8e1773e34d24a"} +{"id": "train_07342", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "evaluation_metrics", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e952ba9111c4861d9862c03182790bcf4ce94cb"} +{"id": "train_07343", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "auditability", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "777a20f4cf4e4b4c212d7cf116e3cf382258502a"} +{"id": "train_07344", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "reproducibility", "documentation", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2bd3748165dfa64312ee77f2819a3c58ffe7143"} +{"id": "train_07345", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "681f926274a12f5e7eb9ae9566c8ce22c7b86adc"} +{"id": "train_07346", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e9b0dfafbba972656b3e207d71f3aa1f6347d75"} +{"id": "train_07347", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd9fd69273a26ef37e26d2b222c7c440401d64a6"} +{"id": "train_07348", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "documentation", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fcac065030dfdb9990160e96bc2ab9a1e285320"} +{"id": "train_07349", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "governance", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b20ed7e02913654823ddfce90bf117390e8c6af6"} +{"id": "train_07350", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "governance", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30e8a21ec10839b32a979bd9d2052ebd69e41450"} +{"id": "train_07351", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "governance", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15c103ba376755d289d36fc591c533f4dcc1bc49"} +{"id": "train_07352", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tests_are_truth", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce556a324952440a812f8a16245d24c5ba8a15c9"} +{"id": "train_07353", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56bd5ec8d394785fe4413900efeb81309c7a773f"} +{"id": "train_07354", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1535c4fe7009fc7cd8429623b3180d4a3bca7033"} +{"id": "train_07355", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a1238a2284adfafd106905f3a1e469042eed720"} +{"id": "train_07356", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "documentation", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60875525e86e4dd2aaa9bc6f18d9d999b6e9a760"} +{"id": "train_07357", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "governance", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81e6498a3a2b91a4dd98fe58fce83ac1321b1627"} +{"id": "train_07358", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tooling", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e761b08786917b7a5ba908bbae30c2b23b3ab12c"} +{"id": "train_07359", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "governance", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a4fe59eac2d6572b993e72f8cd96eb94d7c19d0"} +{"id": "train_07360", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "governance", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35c3f258213264d0fb9a8e22d9e1ab5f0efc419e"} +{"id": "train_07361", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "security_gates", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d293085a3144b9239f48bb8fb6946ef596ed9b2"} +{"id": "train_07362", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e29d592cd98b5a780fc1209206fa9de2053ac069"} +{"id": "train_07363", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "repo_scale_reasoning", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66cdd52c9a4a51f169cb488f51834214af7f342e"} +{"id": "train_07364", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d57b5392c1611bc37b5673825f1c37ab4e54922"} +{"id": "train_07365", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fceae268ce611d1920122539cd25805ad8f2720"} +{"id": "train_07366", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1d9c9587ba11a28b20cea8d432074c1015803c1"} +{"id": "train_07367", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "security_gates", "tooling", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd04c65643dbb0b8ccdbc0200279f05d6e77943d"} +{"id": "train_07368", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "documentation", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18240ab272b8adcfc674a14931d3bae88fbc25ad"} +{"id": "train_07369", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tooling", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc62b09014f22a12a3a2bb5343f72c4f468482b0"} +{"id": "train_07370", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6231a0e74670fa4faeae93e286d23d238ab483fd"} +{"id": "train_07371", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15c103ba376755d289d36fc591c533f4dcc1bc49"} +{"id": "train_07372", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "tests_are_truth", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80731ab7f18d7dde875b09820bc850aec3348212"} +{"id": "train_07373", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cca97d51283b3989a800d4f9a16fcd96ad6a1e88"} +{"id": "train_07374", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bd425c5c98461d553ebad838e25fcadcdf04145"} +{"id": "train_07375", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71a50d1b52c993475fb517dcb7cfe1e4f91ae876"} +{"id": "train_07376", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6937b78c23dfc6d52c3550e37c6cd59200224028"} +{"id": "train_07377", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "documentation", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7891b5a1bde164d1461e01dd514ae814d76aabf"} +{"id": "train_07378", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "752a044ac99ed40945fb0b7ffc9c2ca3160f5bc0"} +{"id": "train_07379", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c071297246d610823e91dc2a5ff15bb6ab99832"} +{"id": "train_07380", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "ci_integration", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31a01e4d3d3445ae6725b40fbca318d695d87422"} +{"id": "train_07381", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa936397070914f50a284a14749ab2495e15795d"} +{"id": "train_07382", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "522c8569dc035f74742e008a1c5ee90c4e45e676"} +{"id": "train_07383", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07088dda478536294935a041f458d74bd6d7faff"} +{"id": "train_07384", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6219950162418951b7aa8033705cbf8855baf72e"} +{"id": "train_07385", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e869699479845e0b9483e8e505dd6d788b670b9"} +{"id": "train_07386", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "documentation", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ba841c79c50b81843770e86753c54a7bdba18d3"} +{"id": "train_07387", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fba2b8bc2dc7ebb8790ffe116a70406407430e99"} +{"id": "train_07388", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b2c8c8c0b50771e8f7423df56e5fd919085820a"} +{"id": "train_07389", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2c9b9fc8b40ed8d11adc6d8b5bfdbe2e986b73a"} +{"id": "train_07390", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "security_gates", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ae225aae0f89f62b61f00840d5b835f4f46fd8e"} +{"id": "train_07391", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "ci_integration", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d1906a9c76b1246e84e63f7dc69408dd14eb77e"} +{"id": "train_07392", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ca8ee6a4787745cf1693955ac040ca507148770"} +{"id": "train_07393", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7d91782b768e79ad9c41590c0fe55dfd4de6f71"} +{"id": "train_07394", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "governance", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bcdfc1d58fe6db1321835baaf7b22df9ed1ebcd"} +{"id": "train_07395", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d20feb0cbd729e8de3ac332d3f13c7a82095f603"} +{"id": "train_07396", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "documentation", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07b9c501e82548571c53a2ce8eafa22f6e042866"} +{"id": "train_07397", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da00b24e79a100621a0113b76de4b1ed7e6daa8c"} +{"id": "train_07398", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tests_are_truth", "documentation", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcc0adc0e872f933044f7f4c96b9eddff947815a"} +{"id": "train_07399", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "143294b4747a05dcb794fddd70e34d27239cae8d"} +{"id": "train_07400", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ff3f773892dacb59b5fa5e0c0a87ab2d1bd51d6"} +{"id": "train_07401", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "documentation", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a944281425809492f1cf167373156382b931abe6"} +{"id": "train_07402", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "auditability", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4d413f4e61b5d8b8de792bbbfbf1eda115996ad"} +{"id": "train_07403", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10eb1962f692fb36c12c86d774fa9aca2df06e0e"} +{"id": "train_07404", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "governance", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e656538733914f1326033da33fbbc92ae9247caf"} +{"id": "train_07405", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "ci_integration", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc568fd7e945add35b5ad8842bbf24979e5b69b6"} +{"id": "train_07406", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd8ddf4ea5f29278ea5267aa5fe70de08e7da3c5"} +{"id": "train_07407", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "repo_scale_reasoning", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cd81a5b60b59ddc08cc99e2cd9778e8a9ea8442"} +{"id": "train_07408", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "documentation", "governance", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9df6e1baa33ba82d0956c83e57e3cfb674efff88"} +{"id": "train_07409", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "documentation", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa286d11b86730a34e3879439a8bd182a41d031f"} +{"id": "train_07410", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "documentation", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1bb897aadc1e9059c721e8a82884ede593d577e8"} +{"id": "train_07411", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff167e2dd89774f30df63247c7a06cd505951765"} +{"id": "train_07412", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0aae23422740a5ce0bfd505a95aa7ccf2e5f6b9"} +{"id": "train_07413", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67c9df4f08cac10ef6ac24f0afcca7934580184b"} +{"id": "train_07414", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "documentation", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1e8b8166b9e7a03fc87bcc24c99af3c414038fa"} +{"id": "train_07415", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "ci_integration", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6b33e3e336033e20bb2b61a35ba4a905382e4fb"} +{"id": "train_07416", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71e29496cdba2867c2fe9ca0481ce2b9b06306b1"} +{"id": "train_07417", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "auditability", "governance", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe3f6ed84e5dd71ca863a3dcf06155ce12faf239"} +{"id": "train_07418", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "auditability", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60010612d9807e7c005b9849691a4a6d94341ada"} +{"id": "train_07419", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51024c9a9cc844fbdb0c622781db61da51e14fa5"} +{"id": "train_07420", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f24f273240087bbdb31b7523123e4dad094c49a"} +{"id": "train_07421", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "ci_integration", "auditability", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3259e656c0c3013c1944063645faa44979e26aa9"} +{"id": "train_07422", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "tests_are_truth", "governance", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "193a2a7aeffff0b573cf72b4fc592209c36eb2dd"} +{"id": "train_07423", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "documentation", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e04fda6897b9e31c7810ee017e71a50f1e20e4e"} +{"id": "train_07424", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8c1e72a2a4a2aed55324ef6e385a210b6e7e564"} +{"id": "train_07425", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "security_gates", "ci_integration", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34c3a7889a98b1e853a393b5859488658c3605bd"} +{"id": "train_07426", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab75663ac324d21f43c5a5e885fb9d8ae99375ba"} +{"id": "train_07427", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71eb5aec323a5569030165330738e5845b52cd21"} +{"id": "train_07428", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb7d66befaa9163cd0e6df596a9375f049953c3b"} +{"id": "train_07429", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tests_are_truth", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36ccf8d160aa32177598b80e43ede6e05e6543cc"} +{"id": "train_07430", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45f1a6e59464a514081dba805fd224edbaef7e38"} +{"id": "train_07431", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tooling", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a822c275e3e8807f24808819aaff8c76856d43aa"} +{"id": "train_07432", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3093956d9215923887ba740f2d40b50acf9fd7e"} +{"id": "train_07433", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "governance", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de2aca3aba0cf05f05db71313928d34fd5143ccb"} +{"id": "train_07434", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "275acc72a34e7febdb8fd6784f74ec16d4da1a15"} +{"id": "train_07435", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9eea7e74edbbfd109b09715c98bb566e6c686922"} +{"id": "train_07436", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d745f842aaff0cb53bb09e8335cd46e025ae338"} +{"id": "train_07437", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5abefee442e95ab905415390584114acf0b6a24b"} +{"id": "train_07438", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2a479bfc5d1d5845fbd89e46f0c96c283b31d4d"} +{"id": "train_07439", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ad836ed42b6dc637d8211ec3a198a01ff54ce0e"} +{"id": "train_07440", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d93554c43c9958d025fb4b89cf0a3819bfc1d1d"} +{"id": "train_07441", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ff50352b9109d1e30241c822384c6859aacc294"} +{"id": "train_07442", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ac14f013f59755f329996b10f1159a6782d35ce"} +{"id": "train_07443", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f4c2ce6f327a8aae1e111a83fca47c5f1cad05f"} +{"id": "train_07444", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64c5c602a2a247590acd406423b22de5204e51e0"} +{"id": "train_07445", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5a19847d077e9dae2f0fb63f47624fe4721405c"} +{"id": "train_07446", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "security_gates", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ff21702c95cb9f4731e3152991395ca8b9ce445"} +{"id": "train_07447", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef985048ee27a820a8a2856c0e9db98b9ded104"} +{"id": "train_07448", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "documentation", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffd06b0f95d0a277999858e8eea60afe2dc30c0b"} +{"id": "train_07449", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tooling", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d8deba48096832be99b16ec86d3eccc61789732"} +{"id": "train_07450", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d26c4399a52eb6878cfd2ee399d9695365c7535"} +{"id": "train_07451", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "tests_are_truth", "documentation", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8db074de0d702d669228656a5406d34a0e626dc3"} +{"id": "train_07452", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "365f16a31a500ba8c55b50f16019dd6e10ad3808"} +{"id": "train_07453", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "511bb8241930aeb214d399bd67edceb5e2330d15"} +{"id": "train_07454", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "reproducibility", "governance", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a200c501fb823f68d1300f2ef1219f5fcfd4b9e"} +{"id": "train_07455", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "526211db4c366083cd4b243809f19c5d7976c722"} +{"id": "train_07456", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8444e6e6e9d916333b9ceb22fdcbbe8d8cb90e94"} +{"id": "train_07457", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5869ebbfbab195867ea70300fd8bfa4080217117"} +{"id": "train_07458", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "documentation", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e8fd12f7909a302748d10c4f8f8c807e4093a2d"} +{"id": "train_07459", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afb3e4987f840517385ea0a1ced148400128c0fb"} +{"id": "train_07460", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "documentation", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e487d0618e89ec583486a6f48c8f59734f5defef"} +{"id": "train_07461", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9efa1a9a772332e130f6d88ea946cfb361c36f36"} +{"id": "train_07462", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tooling", "reproducibility", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b92cf36fcabf5052f9e71aca9b09f9cc58c21c2a"} +{"id": "train_07463", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "ci_integration", "security_gates", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ac43057c2e2da09bd5c79a07de653aa480952f3"} +{"id": "train_07464", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f84af4a6996eeb4976689b9186fc2407c97e196f"} +{"id": "train_07465", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f566b71874b6ba09033e7061cb5de2a1ce2d00ed"} +{"id": "train_07466", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bac47152894583a8a9dd7c9b91a510fb621ea5fa"} +{"id": "train_07467", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bd73341c693fd9fb6b0945c86544dfc02324a1e"} +{"id": "train_07468", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f66903cbb736fb5776799dbb1c75cd651d57dddd"} +{"id": "train_07469", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b44c09964d2e7c43c067f4754170e145f459dd5"} +{"id": "train_07470", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "security_gates", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3e4f59ea7445f28c18fb6f8c570df406aaee4ad"} +{"id": "train_07471", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "documentation", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfc67273cdf15e87aa9c0f8c09559c8579721c3b"} +{"id": "train_07472", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tooling", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8bb467fe16f0393caf8096ce3c4346b105140ab"} +{"id": "train_07473", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f048f79c5f7cc1236f937c415b281c80a3668554"} +{"id": "train_07474", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c424773423074c1467bbf1a6e6b9ef3461dc4057"} +{"id": "train_07475", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6150597f06bc7737bc8eb183f2238d0f3ae59b85"} +{"id": "train_07476", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e47e3c8b282b6bbfcd3d8368d3b1326629421fc0"} +{"id": "train_07477", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d86d38ef2992abaa6e78359d0475089870cecf97"} +{"id": "train_07478", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bada5367d321c4e6de66a66dac06afddd7441ac5"} +{"id": "train_07479", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30e01f09b241438fcdc05c168a3c03cefdfa8bb8"} +{"id": "train_07480", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0301e70c8786ed17db0265c4ef831a09c8d21c2b"} +{"id": "train_07481", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a5d828961aa819570aa57a0b81670cb85d87e43"} +{"id": "train_07482", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cbdbb46984c0489627834d9b17d0adfefb30900"} +{"id": "train_07483", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d01a3b8200df41c4c6b7d1cc0fbb0ad39a11019c"} +{"id": "train_07484", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ce9b0b2e70262a0c702c4b304da793e5615760e"} +{"id": "train_07485", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "documentation", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a214d75b27326e37a45e963669b6ddfe43d2c72"} +{"id": "train_07486", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46955bb912c535f2d4b75fe67e75c492aed62aea"} +{"id": "train_07487", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "evaluation_metrics", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bf52e9f8d52ed2b8f6a67f57d44676ae76b0d2d"} +{"id": "train_07488", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "436d464c6c579711ba3f2c530ec5bcd279b8dc1d"} +{"id": "train_07489", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd67efb28b1a331d55f066c40e6c36b370d818e0"} +{"id": "train_07490", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aabf985f25b63adcf4442f1386d13663ef03fc6b"} +{"id": "train_07491", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89a4c6fc21ddd66f3a91f66063181d5ceb572ace"} +{"id": "train_07492", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b44c09964d2e7c43c067f4754170e145f459dd5"} +{"id": "train_07493", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "reproducibility", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c8775b9c1806823423bbaa32fb989766933d476"} +{"id": "train_07494", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "319f22549bf364ee3424ac6e30a5c177e5ad5eb4"} +{"id": "train_07495", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "governance", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19cca3f8ce6784c1dc734f241dc50a3ef81d8b93"} +{"id": "train_07496", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a80586b1ae8934e275fb65c690364e34e1cb903e"} +{"id": "train_07497", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e73f97b10712c051bdb2949a52721e850f1d14e6"} +{"id": "train_07498", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "governance", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be809b59bb187545c239f812d489df5e1414dcc9"} +{"id": "train_07499", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "839ed65cdbc1c25dbcedf60101d9b1ed87f68857"} +{"id": "train_07500", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8f0510fd4fc3531e1a1bb757c7b81ca251089b0"} +{"id": "train_07501", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "632509e04dd02132c0c182e54a53856b0284bbc9"} +{"id": "train_07502", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "governance", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "567f58a4aa5ca12052118d80b977dc2778fe1d4d"} +{"id": "train_07503", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03416376969c0532b5f8ab78fdcc2bbb3fde8301"} +{"id": "train_07504", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c95b593fc1d6dcafcdb6740e8c9f95669cda67f"} +{"id": "train_07505", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "ci_integration", "security_gates", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73ba525496d7f5e7d21da1092dc4e0c5f33746e1"} +{"id": "train_07506", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a9be9cc0999548d743aa97b982d2a766f7299cf"} +{"id": "train_07507", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b47895c32b28c317ac8eca42a90e52d9099e67f0"} +{"id": "train_07508", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf79ede3bdd0b72e0bce0bc6ec84e14c49a119ea"} +{"id": "train_07509", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "governance", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecff726db656e92304e7099ac8205a982a345f32"} +{"id": "train_07510", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7eeb2ec63c426fcd69cf73bd0ca9fbfe760e4755"} +{"id": "train_07511", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "documentation", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ef523b4b30c563165e0b9cec9f1564f71c9ef74"} +{"id": "train_07512", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b0a9b50aaba5479048b4a9651d7657866489bfb"} +{"id": "train_07513", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tooling", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ecda87ed2bb168dc4a494406239b6bc7477d48a"} +{"id": "train_07514", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tooling", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe2c05e78551f41c2abc81ef4a9c2cb687f53ba0"} +{"id": "train_07515", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff88bd9574b860c2fa750e40753351a17f12ea86"} +{"id": "train_07516", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9eea7e74edbbfd109b09715c98bb566e6c686922"} +{"id": "train_07517", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "636b0f02181bc27da901c488dd37228f820e6078"} +{"id": "train_07518", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "evaluation_metrics", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4dbb95994d891b93872dc7bceb42511c467d665f"} +{"id": "train_07519", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb1d3a9c8834835f473dd7701edec6e65eea9a69"} +{"id": "train_07520", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afd6a0cbabcef277fd9eddaf0861b835ae7b1cd6"} +{"id": "train_07521", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "documentation", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7fb2b2c245c6106485f86af8c6aecf05a9d2016"} +{"id": "train_07522", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7eb98948f6cff06958fa5730d174699aa98cfe58"} +{"id": "train_07523", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "add1af7962101545f19a960df254000d303f7531"} +{"id": "train_07524", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "014415df47bc9532b4b1792d33a1c263ca15827c"} +{"id": "train_07525", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73f60dcf6ea236f13915295e4e0f4cd73759ba66"} +{"id": "train_07526", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "tooling", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcfdb8d7ab08340fec469705774f5564c9389778"} +{"id": "train_07527", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27bfd42441932118b3af26ff4850bf4a81482580"} +{"id": "train_07528", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4e8b8e8ae74c7d471121e3f4a9fc8c00fed8fba"} +{"id": "train_07529", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "277dca203bf1c3e829e9e5411b487d90a968a52c"} +{"id": "train_07530", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "355cc3fe6f077e6ab21eb44279144bd47b631f03"} +{"id": "train_07531", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6a832dbc2ee989096f3ec71b1d95518efc54cd2"} +{"id": "train_07532", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12bd84286cc50f6ae173f0987decfc8299ee9e23"} +{"id": "train_07533", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8afb16a068a071018c4898801f511abe8ddf4d16"} +{"id": "train_07534", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "governance", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3d02dfc330a739aa56e7d438c43ca3ee9113e81"} +{"id": "train_07535", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "security_gates", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccb9b52ff5786742793388ddff563c44d92e8b2a"} +{"id": "train_07536", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "636c0ce3c826ec6a155ed8b66426c50fcf9fb6b5"} +{"id": "train_07537", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09ee32c8858101348849d4514982b8e919dc2ec1"} +{"id": "train_07538", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "governance", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e424cccc5cf3eb7212823c62a17cf96c81c63071"} +{"id": "train_07539", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "auditability", "ci_integration", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2b7dd443b9fea4ec0e28fdb6c91b5449dc72206"} +{"id": "train_07540", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e103b385f326008cf7ab09d5ed3517430d1e23f6"} +{"id": "train_07541", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73ef181bc7b83a46da2c651c5fe55685a7503fa7"} +{"id": "train_07542", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "722c2d6a88f933f47fa5bd9834d684015d654e9c"} +{"id": "train_07543", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "340c31399833acf17249d7c61dbf750dd7672df5"} +{"id": "train_07544", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "reproducibility", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9ed7923b0b85011101e00a376d029e220f19d2d"} +{"id": "train_07545", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "ci_integration", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "679044c575fbcb6e4b462cb5919d4230cd61153a"} +{"id": "train_07546", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7178dd7bbc96a732ba037defc3eea06e01ad78bd"} +{"id": "train_07547", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "documentation", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddff4f14c989a00c25bac6aaf2243bfb6ef74bcc"} +{"id": "train_07548", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "163222f58d5143a89b2f6c26defe01447345f302"} +{"id": "train_07549", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "reproducibility", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0690c7c6aba656b7c6316df3ac973a5ad3fb1ec3"} +{"id": "train_07550", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "412ffaa3f6f0d6f1e6794bbe91f1f4cbee88ecd9"} +{"id": "train_07551", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e0b0c507181b48c56086634255dfbc3fedc439b"} +{"id": "train_07552", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c891defad4c09170056b2348d6fbb6b80616f5e"} +{"id": "train_07553", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1b4664aeae25e2211874d7252b4a000f5a8eca5"} +{"id": "train_07554", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e47e5ba9dc4050cc0110aa9e07f62c170a72b444"} +{"id": "train_07555", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "369f8476612a04791f506b5300e7041e2f749506"} +{"id": "train_07556", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ead198207e789ecc83728603d43df1ac4c33fb1"} +{"id": "train_07557", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "documentation", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e68ef07af100a939f2c41ab3c1d41a5a1f81977"} +{"id": "train_07558", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fda2bd2ee98ae683ba6ff407324a1ea2a7db0cc"} +{"id": "train_07559", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e72dfed53629201d4721cb086a85ae71be4e2493"} +{"id": "train_07560", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "security_gates", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bccf9052f4fd0a5bc0127ac0225e2f0f71e6fd78"} +{"id": "train_07561", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44e64ab400e8d784a92637a92a4e2d6417889fd9"} +{"id": "train_07562", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "421d79bae2c94d4926e0b9c135453818a3c7cf95"} +{"id": "train_07563", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea8a235f197e0ab16d75ca79821a794019b6423c"} +{"id": "train_07564", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "documentation", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "480aaea3325a2a5d379fb0ef2a58551f820febc2"} +{"id": "train_07565", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90415e0ce59005266deeb6826cc2683ff75ee16d"} +{"id": "train_07566", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8706c3596140063441bc438782a867bbcc66048a"} +{"id": "train_07567", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34941ab4efdb8b5fe48d42dfc74e0402457f6ae3"} +{"id": "train_07568", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "tests_are_truth", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d32eb182f1365feb0ddd24fa330c220917cc46eb"} +{"id": "train_07569", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c5cc823c6893235f796cc8a06cbe9e477912310"} +{"id": "train_07570", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91c2150330209edc3489df8a957cdfdd82826e77"} +{"id": "train_07571", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "reproducibility", "tooling", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab558ed5ca364becb7522f58f584b3f85e6f9266"} +{"id": "train_07572", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "ci_integration", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bda9a0919de3a90887adccd9eef80fd02997eac"} +{"id": "train_07573", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "documentation", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5acd8ccc6190d420f139704c62226622694bba0"} +{"id": "train_07574", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "32723a0ca2376f8e66d890b2d1cb75430a7558bf"} +{"id": "train_07575", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "security_gates", "documentation", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1554ad72b1c4012cb316acb1486de92d6badfcc"} +{"id": "train_07576", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8762852519ec5ff27bda31e9187493f6916650e"} +{"id": "train_07577", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d01799615b56257257686aff5d1086dd3a60fc50"} +{"id": "train_07578", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "reproducibility", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6ac51dee9349a5815c4fba83cd4912a17b6a0e6"} +{"id": "train_07579", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c1c4fedfcb0932767db0fe811d49ea5a1f810e7"} +{"id": "train_07580", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8baa4e53c8148c384689e9fdaecb6524bd86c632"} +{"id": "train_07581", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b70f20d5b269464bfc9c7141f378ebe184808672"} +{"id": "train_07582", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85e11ad646f178713dd7efa0a0ce46293c0a5ab8"} +{"id": "train_07583", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c469f5469aeb28a4c53e90b0e02a948649aa5b4"} +{"id": "train_07584", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "security_gates", "documentation", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7af38e29e43402c22c16098393cfc68c2582a37"} +{"id": "train_07585", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95fcce900e805b519c48bba2234fbf98ae49e5e3"} +{"id": "train_07586", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4981d25813f9437fc1dd88c049a5f1ca4aaec9a"} +{"id": "train_07587", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6899f4666f584d531f944a87abad7e105ec43b37"} +{"id": "train_07588", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4fd7eeb252152489a64503e952040f4523d1795"} +{"id": "train_07589", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46862f5cfd3eaf4658ffb65cb4f94520fb939e7b"} +{"id": "train_07590", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25969f8acbabb35f5fd10a3e31ecc86c93e0be4e"} +{"id": "train_07591", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d93ae20cf760bbc74eb586b84982a22cf193ee8"} +{"id": "train_07592", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dccce11774c8740c0c2445ee7f23b41ea8f6fe7"} +{"id": "train_07593", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcb4f3d795176c651478b644755fb48e33d4d7d1"} +{"id": "train_07594", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43b246963c9f860c86c8c618af834b504b172725"} +{"id": "train_07595", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "036d454000db08718c807de2ba6e9d87a017ae01"} +{"id": "train_07596", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "documentation", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46c6fbec484e107e023a0eb59c01a9eb0ffec441"} +{"id": "train_07597", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d9a62b37c1c9f2ffa9d6a702ee7232ce74904d3"} +{"id": "train_07598", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95831d0943ed9c3249ddaeb518e669091b7d6e2d"} +{"id": "train_07599", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c28ce615447d948bf6233ea2be161037528c0d7"} +{"id": "train_07600", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "documentation", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8a3b6a08f942e90ce72a10a00b5232624efea48"} +{"id": "train_07601", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09d65e331e02f647b74923a927893382fe191844"} +{"id": "train_07602", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "governance", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93c2c7b761bd4d1df084641f1d5904c4c9b3804d"} +{"id": "train_07603", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "tests_are_truth", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f18ce8adf3f3a870f41706dc054399bcd6e89626"} +{"id": "train_07604", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb2cc10b9b7f3c7b4c2e68676f87cfd8df56bf8c"} +{"id": "train_07605", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "repo_scale_reasoning", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7deeafb6097c3bb3532dafaa9336412e23e1705"} +{"id": "train_07606", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcfb392fc468a370d25d640413619433d01122b5"} +{"id": "train_07607", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "security_gates", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b77f35f00184b8293353eddf7da2742a1e796a2c"} +{"id": "train_07608", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c9a4b5d42892b8b9a07a2e2bca4dc565f8ad94f"} +{"id": "train_07609", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "documentation", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a186e87ba5b6351d53019a599bc1d7626f1b6f93"} +{"id": "train_07610", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6af45377eadaac268716343f3db6393b8e31a338"} +{"id": "train_07611", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d6304a6f536583ab57b342a59cf5a2f2feac519"} +{"id": "train_07612", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c5d91a3b6ec14ae2a776ee594a5ade492dc283a"} +{"id": "train_07613", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae530bab0cfb6ca46bd936307fbdd06d2c1aef05"} +{"id": "train_07614", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69a4c1f55f09cc32e0226d7b9be9472783a514c4"} +{"id": "train_07615", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "documentation", "auditability", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5c2be0c741cbff3fa462a8d39ab8bc1e1a24379"} +{"id": "train_07616", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "ci_integration", "governance", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "193f9df08bac464b03d79b4bb2e6d66dfcec981f"} +{"id": "train_07617", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9db6ddff0fe041263ad850dfbee2689d8bd59258"} +{"id": "train_07618", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0593b036da998ceb3cdf6f38229f3550b6522db"} +{"id": "train_07619", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52e996297ad1bc0f1904fb8519091147315eb4c4"} +{"id": "train_07620", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "governance", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5561db8a56fd1113096a8433ea2dec1260b86088"} +{"id": "train_07621", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "adaacdb406c7b90253e4d477ebdd7eceac219de3"} +{"id": "train_07622", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "ci_integration", "documentation", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78a240c619e759d8601f264b06d47066b46bcb3c"} +{"id": "train_07623", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cff8ead94eec108b63238a06716d6c73722cbae0"} +{"id": "train_07624", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b36edfc37f0408af561504af876afdcf45f59c2"} +{"id": "train_07625", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3c2dd4ed8e0e3f961c5a0edfda2e19e2710acc2"} +{"id": "train_07626", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f9e6c557ffd1b7fe4db5deb19f7ddfc6c882705"} +{"id": "train_07627", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f48fb7286c9c363cd1e851eec497f7acdb5fb29"} +{"id": "train_07628", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52e37b6ee064d3d9c0334f7ea6ed04719b073fb9"} +{"id": "train_07629", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "auditability", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f86a160aa999d9aa5554e0cddce18f1bab9c902"} +{"id": "train_07630", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e48abd6c5fb4bb0c1d7e98efa041c22ac0ec01a8"} +{"id": "train_07631", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f35d088355f891fa3b203f4278ae680a9f165296"} +{"id": "train_07632", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "evaluation_metrics", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "562289edea085b49f241c78b07f3f75016739dfc"} +{"id": "train_07633", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "646f834054ce7d2605f473272297922a8b4ebd74"} +{"id": "train_07634", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "documentation", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "401210cb69046b5c8791915354e889b20ca9eeb4"} +{"id": "train_07635", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "ci_integration", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a082729be68cb51fc7ea4e64b782131b850082ab"} +{"id": "train_07636", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "059221545e169fc00f8364edd38cbb60b9f94209"} +{"id": "train_07637", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tests_are_truth", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7606cf7bb103a4abad41ad28d6b6ec1d6b58a3cd"} +{"id": "train_07638", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "tooling", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b79531042cf95a55db5ad90bd49d8a642aed2686"} +{"id": "train_07639", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d815a0d2889aba0b84d17bf6c67eb3861a602b71"} +{"id": "train_07640", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3aa0b013654b7b12ff242d02215de118e0c28e8b"} +{"id": "train_07641", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3feb1389744c37b074c32fc7d487f0a853081a7"} +{"id": "train_07642", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe1a214f61e5bd01a9e6835866337ca00c8e499c"} +{"id": "train_07643", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "tests_are_truth", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cc03e9fe964251fbfdc3f60f585cb41c119167f"} +{"id": "train_07644", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e061d24196a68116ef228e3b517ed9b476cf1037"} +{"id": "train_07645", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "reproducibility", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3feb1389744c37b074c32fc7d487f0a853081a7"} +{"id": "train_07646", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7f536c8d99c8891b546a248bfe540e96af8a8e4"} +{"id": "train_07647", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4c22d4f8d017f5c5b50d08cc1275ed82153cae2"} +{"id": "train_07648", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c950932a38c53eae54ddfb02da60eff4d0212e5"} +{"id": "train_07649", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "security_gates", "tooling", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "834d95543024ff4c0f73b1e2ebc6fd40b8d9d60f"} +{"id": "train_07650", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c11dfbc7b389a87d89eeefaba429e1f0e60a616e"} +{"id": "train_07651", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3007dfbedafe53c441ad48dcc30280b5bf883ed"} +{"id": "train_07652", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b22cd9822ae6044e285f515e01a081dafcc99681"} +{"id": "train_07653", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e15f58c06ba9a6bbf593975bb659de1643f9d30f"} +{"id": "train_07654", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "governance", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "415e4c9ae19e12a7a94ec631fd2a63de2fa4aaf7"} +{"id": "train_07655", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc002777883df6f63bd1d1efe50aa0348c651427"} +{"id": "train_07656", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d991c6ea4bcd056bc06335b7264f4f87465d1e6"} +{"id": "train_07657", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "ci_integration", "security_gates", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9827c06b35bd4161421a911040357fac51ca557"} +{"id": "train_07658", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd7faf1113c7bb0842672471e378c2c4af23302b"} +{"id": "train_07659", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "289b5880b6b22d660a170842a93560694eaa814d"} +{"id": "train_07660", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "cost_latency_tradeoffs", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c98a95125bce146f711c022cbc615b31aa021d6"} +{"id": "train_07661", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f0462e06919ddde8923fba129857880e9b37fb4"} +{"id": "train_07662", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d9bd54e775dde0812f6720662b6a1e655c5d35d"} +{"id": "train_07663", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de442dec91cb51603955f87ea10a9c8d9e817752"} +{"id": "train_07664", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62d64fe2ea84e298c2e960757cac7fff18c5465b"} +{"id": "train_07665", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17dbea969d3744c642c61138421d7124d29fe950"} +{"id": "train_07666", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5aae1084e9c90e87a94dc174a449882a7bb86eee"} +{"id": "train_07667", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79723633f3177d9a6d70bcf9952a97a7ca0841f3"} +{"id": "train_07668", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e757ab862860eb315a5382ae29a6c00f3783ab97"} +{"id": "train_07669", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0673bf913c7e3be79130008358a834a605693b39"} +{"id": "train_07670", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfa9ad1db229dc876b5430bc17b4ecd6cd44b379"} +{"id": "train_07671", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58dae66998295f3b6909fe46d571715252c239b0"} +{"id": "train_07672", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "evaluation_metrics", "governance", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be1828ff4c78b4450fd21f1a7b9a3f5b7ecac8fb"} +{"id": "train_07673", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e68ef07af100a939f2c41ab3c1d41a5a1f81977"} +{"id": "train_07674", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "953c46c65ec0c0aaba5cc3b6dc608a22f369f008"} +{"id": "train_07675", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4c4889fa2f30e9616f65a518a72dda0f3dd7d73"} +{"id": "train_07676", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65dddcded88f699e01ddf12aa6d1157ba5c110a1"} +{"id": "train_07677", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f03cb5a0bca0e8872310903a6c948d64ca40b945"} +{"id": "train_07678", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62eccf0ee356510f1999438aace320aefe032e5e"} +{"id": "train_07679", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0540dc988a17fd8216f5159291da13bed9ff959"} +{"id": "train_07680", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "auditability", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fba9d0f0ba1e6c08baf7b6c28661fc2a8dee96d5"} +{"id": "train_07681", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "documentation", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1960f73bcf54bc241d619d8b48cd23bbeba15879"} +{"id": "train_07682", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "reproducibility", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67f2014e44227089797cd9e6612f3838728415fd"} +{"id": "train_07683", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2282211f84db8b9c589e17e044cd75e3280690ab"} +{"id": "train_07684", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47d96750360083eb693a6908569ce2c4206bc73b"} +{"id": "train_07685", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "security_gates", "documentation", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45a744e6246cd97062356caca90773660723f104"} +{"id": "train_07686", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5e7cf0667b14475c356ef48dba69a91ad6ef061"} +{"id": "train_07687", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25d5186c72aefc35f08425caa413f1c41b2678e0"} +{"id": "train_07688", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "documentation", "auditability", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19beb05cbc89a2b3b0b8a2503ba0e06e0f4b92f8"} +{"id": "train_07689", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f41aa070cb93b8e033a8b533f8fb3cf8b577de9"} +{"id": "train_07690", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b5a38e73dc5d5c3f592b2155ec12eb20fc9572c"} +{"id": "train_07691", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5cba41f1ae4750d8768cd0619dd924dc13f10ef"} +{"id": "train_07692", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "ci_integration", "evaluation_metrics", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f7b90c0356a2b22a53b78f0d6cb48f314faf9d2"} +{"id": "train_07693", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae8a40f6e80f927f71850df45e1c105541a3a2cf"} +{"id": "train_07694", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "security_gates", "tooling", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d022aa125557194f8b9d7f4a378903a87f1d4ef"} +{"id": "train_07695", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3947672bba12a98562d5b63de2e16590cd462746"} +{"id": "train_07696", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "security_gates", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4778fcf5c0903719b9d37754b23fa0f50da31db"} +{"id": "train_07697", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "security_gates", "tooling", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ad94c56f8f25543c8c097aab0a020d5720e70fb"} +{"id": "train_07698", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63465dcff70f5b810c491b426563577d3dbe0dd3"} +{"id": "train_07699", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tooling", "governance", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cc07c92c3c36b4b929d19b203a0ec02a378ff84"} +{"id": "train_07700", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "documentation", "auditability", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d222fd7b4585ea90108a442f925b8a60d660ee0"} +{"id": "train_07701", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d704582aa9bad0580d10821abc317ce11ce71c34"} +{"id": "train_07702", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8cf486aabe812267274254b29b24258e8911b8da"} +{"id": "train_07703", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d39e2cb1a32b3229c877982178378cc7c1a011eb"} +{"id": "train_07704", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b95b94290afaa00f27e7313edf4e481fba9d46e8"} +{"id": "train_07705", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8444e6e6e9d916333b9ceb22fdcbbe8d8cb90e94"} +{"id": "train_07706", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "ci_integration", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4532bcd9c8fa71ca73d31fed3ba4fa85403f9025"} +{"id": "train_07707", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b3c88d0ad15d1db767cdb7894e2540a64ea51d5"} +{"id": "train_07708", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e545ad87131d453293ca1888f31b924b01ae7ce3"} +{"id": "train_07709", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1656ea0d376f37f6dd8f353eeb7830cd22ec75ea"} +{"id": "train_07710", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21526d6c11b0c9d273e1e6146ba40760545f508c"} +{"id": "train_07711", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a98920fc5086dd407d52dbb629016804dd767d74"} +{"id": "train_07712", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c27578ecea02725fea88afbc081ba6a06bf64bc"} +{"id": "train_07713", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3f7f509b5ca8afe888b94a13c6118f223c7c6af"} +{"id": "train_07714", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1965a64e1175557b6cc5babb3fbb87050d52ba61"} +{"id": "train_07715", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "auditability", "governance", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2bb68bb34b6749ffdd4922fa2d6c577d4f13d27"} +{"id": "train_07716", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f42b921c5c9d6eed4dc0a89ff3482228ebf5a0e4"} +{"id": "train_07717", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "reproducibility", "governance", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa63c461c806e52eaf29b3d722731e08e0461b5b"} +{"id": "train_07718", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddbd629fa57a6374018fabb8a98e0010b519933e"} +{"id": "train_07719", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "security_gates", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b115a8c9161f348db1525e517dfd6a344f6e780"} +{"id": "train_07720", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c72a0f5afdeda870f8fa27029337f65dd29eb05"} +{"id": "train_07721", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae8411a795d69207c0953bf58cbad41f18c5156c"} +{"id": "train_07722", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4eb64ace1b861bf8d9c387338c9797b0b47a0323"} +{"id": "train_07723", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23502e1f6203d8167ded3e2d25b3a73a0c4b7277"} +{"id": "train_07724", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f60736714a81662d163f2b7d6c313fa1bada285b"} +{"id": "train_07725", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6108e9387fb129dd6f501d63faece6064d2074d7"} +{"id": "train_07726", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f2f36e3dee24967b0766a84ed77d50b91164b88"} +{"id": "train_07727", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "auditability", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67bba64f70a9eed84cab38522d02b7ccd597e82a"} +{"id": "train_07728", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75b728fb9eb9418cb096420b11dca5feed9b0d1a"} +{"id": "train_07729", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff2c967f8def0c9f196dbc1073a1ccf7f249ca0c"} +{"id": "train_07730", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8cb9b7cd3f4dffa79709c982e31ee35baaea471f"} +{"id": "train_07731", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "edc6cfcc4cc8ac7a6fb43bf90a61fa0cdf04e289"} +{"id": "train_07732", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "862852fdd32748ced97fed8b025d85aced2b66a4"} +{"id": "train_07733", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "security_gates", "documentation", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fdf225a25d1b8fad9e4e37126fa2f54738517b2"} +{"id": "train_07734", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tooling", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c3479788d4f730f354d28388301f1c8c064d3af"} +{"id": "train_07735", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "security_gates", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0902f7a4a71655c868da4cc518a6a0affd55e9f9"} +{"id": "train_07736", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6271308fcac03432bffbcd055fe8ac09ad9db738"} +{"id": "train_07737", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e41bfc24fec8770353d44e35f4427cdfbad2dddd"} +{"id": "train_07738", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a12bfde9d1db3a5504960c6826e94fd297024b9"} +{"id": "train_07739", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9199706d032336d99a83682ce320351629a7eeb"} +{"id": "train_07740", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77202d959906043c3389629ab8d5a644c9ac4819"} +{"id": "train_07741", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eec6f43c8bc3cfbeb00df1198068ebcc837ca1d1"} +{"id": "train_07742", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f77a2675034f120563321d705d443daa44bf3ff"} +{"id": "train_07743", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "security_gates", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64f4f42e644512687a9eda4c365eb36556c50e84"} +{"id": "train_07744", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "security_gates", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d269ca425d333a8f33a8de5d98f665153caa1410"} +{"id": "train_07745", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "documentation", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17d2e9262759c5ccb65082fca2317f221296dadc"} +{"id": "train_07746", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ef8bc00d380382f8eda3973437c63196a420bba"} +{"id": "train_07747", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63cbc72a41bed999d8f83ad24d72c51f8313144a"} +{"id": "train_07748", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4d49c794ee288ce5fc92f1aea59ed6814f5e232"} +{"id": "train_07749", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24bde576135aaa2d0045b6b75ef241a892c62701"} +{"id": "train_07750", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18b8a6cd6e976e370749dab790005ddfdb0fc6fb"} +{"id": "train_07751", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87c0f34e08b139fbcfe70defb3a7bc654d26851d"} +{"id": "train_07752", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "tooling", "governance", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "598423a662eaa5e7f0b214e4d701ab5823f97c3e"} +{"id": "train_07753", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tooling", "governance", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78e7debc9fc71d98859ac74ca7daac9f56dedec3"} +{"id": "train_07754", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af2b8c3eb6353ad82adb36b01da55cc52bf6c05d"} +{"id": "train_07755", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "reproducibility", "auditability", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f343dc38bb86575c70c1dceb7410ff77650e520"} +{"id": "train_07756", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40a456c3f154e2e20ca802e880f816bb7b6d8406"} +{"id": "train_07757", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1955b5e5a515980e453051eb8daa3c9d0c79b8d9"} +{"id": "train_07758", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "governance", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bf7cb2087eee20060ce705440c6374d715a31c4"} +{"id": "train_07759", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe4728006520a070c76d2ca936ab12e10407ad54"} +{"id": "train_07760", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebdb70207f1511171e4d7ac67140db0e2d832e6a"} +{"id": "train_07761", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tooling", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "600b6ab32794b15b593506f50d5dada52e1dc47a"} +{"id": "train_07762", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "documentation", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ded5c643a9d74d88f7eac911892c28d7a41fae50"} +{"id": "train_07763", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "auditability", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42294db117eb5ba876939dd3e8994b3275ee80d9"} +{"id": "train_07764", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3598eb91445c31ad563aa91e2cd49d4dc81ae648"} +{"id": "train_07765", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "repo_scale_reasoning", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f1af9e3574f9b4b2f0f60595f4be128e03d397f"} +{"id": "train_07766", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ec5d26aadb0539b2ba3f3304158315418d2ccf4"} +{"id": "train_07767", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "evaluation_metrics", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbbcb56e9146d9307e994fd6af9e038118e5485c"} +{"id": "train_07768", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "tooling", "documentation", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1561d9157c9c4b73e165dd8fd033c30a07fc8477"} +{"id": "train_07769", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c2157e9a829aa17d64480d2dd873778e6ce5306"} +{"id": "train_07770", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5a9f172eef06efeeee0f16e5016ec4c0ab86da2"} +{"id": "train_07771", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "708dec21cad145683fd7718a22a39a0941d22033"} +{"id": "train_07772", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "ci_integration", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9511e8cd50a947ac8c2a2774e748c913e4e908ad"} +{"id": "train_07773", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "reproducibility", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1dae1e145a3c3ad6f65cdaf50ce1efcbb60371ca"} +{"id": "train_07774", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "security_gates", "governance", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "216912256d04bda87a6f5c056010a9d8273a51bd"} +{"id": "train_07775", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tests_are_truth", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c5eb96e266e8041a0d98f78582507b17692ccdd"} +{"id": "train_07776", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b893105be029c638c73c81bb5290b6547f48e11a"} +{"id": "train_07777", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c6e0701baf199027c811968dc2d73591ed2941c"} +{"id": "train_07778", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d52fce3ec1f4ed379c8bbdeee3459a39ccb4392b"} +{"id": "train_07779", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "governance", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49215e6c7fb5a554bd1c072c802424182587d8d8"} +{"id": "train_07780", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "evaluation_metrics", "reproducibility", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4319687a01d7d0888f39381fe904c365e99dd822"} +{"id": "train_07781", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "governance", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b6b58981169ceda19e7fd49777343862d4df4c8"} +{"id": "train_07782", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1e4e45311bd08537dc5efc8906c4a41b6552c96"} +{"id": "train_07783", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "979ebf0dd8b43a62d8bc4b402c6d6ebba64e598d"} +{"id": "train_07784", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23d9a00981708ac705d6c4d68bdc87bfdff4a6f9"} +{"id": "train_07785", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf4cf8715916b2a5c77a51e338357a1e87954541"} +{"id": "train_07786", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7ea0415d92302841858ec39471a24920d245041"} +{"id": "train_07787", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "governance", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c09017d346d9b7c5eeac9844919ad1beb476453"} +{"id": "train_07788", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8aa72b8db1d2abca8e1db8e2203bfc0801f15ebf"} +{"id": "train_07789", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "ci_integration", "documentation", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7da6b426529637e6ba2f345b066c80ca9a9958be"} +{"id": "train_07790", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fd5266835b2d8216280a423275bb418ef1349a8"} +{"id": "train_07791", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "849cc48da4d9279e25d5f495767f44478ad0fb27"} +{"id": "train_07792", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "ci_integration", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bda8e474c3efce994309948d63adc477ab774c92"} +{"id": "train_07793", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9fb5d3d49de36e0fe85384c135d3eb041c10a37"} +{"id": "train_07794", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e359c3a48fc14ba782cfe2fec3ec6c9d381af191"} +{"id": "train_07795", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a5330be73c8ad7c9effa3647b1e6da76d1e66b5"} +{"id": "train_07796", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0552cf7459cc1f45180a21d2c90ec50874e0080"} +{"id": "train_07797", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3516c3dde1c121aa350e854518184f011105bb74"} +{"id": "train_07798", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27b03fd2e5b24f472b46249683ff1fd902434f24"} +{"id": "train_07799", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3dc6f8ec496ecc533992d48da79d1d7af80aff56"} +{"id": "train_07800", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "auditability", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2cdabcc7fa6f01e0a239cda55ea48fa97f15df2"} +{"id": "train_07801", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "documentation", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "788a83f13dd03a183ffcfde8d8859833306a97f1"} +{"id": "train_07802", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cdb2ef3c45d31533c2b2eaefbfbd2b1c1baaac4"} +{"id": "train_07803", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "reproducibility", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b4bcf188c318c939d79b74c0cfec010f125f019"} +{"id": "train_07804", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "780ceaea8e36251df5cfa77123a5fa9065e5c4f9"} +{"id": "train_07805", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "governance", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f14aa7049ef3c204b254ef1445e691912c401a0d"} +{"id": "train_07806", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89b8b3ef719f2c33f645e3aeede2ef4d082abf85"} +{"id": "train_07807", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16f1aae8b61c63bd128fdd67143605c505a3fddf"} +{"id": "train_07808", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tooling", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6830f6d480626b865fe063c298d38aa9581faaa"} +{"id": "train_07809", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "documentation", "tooling", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd377559eecb6115d114fdb8bacb554f214e7b3e"} +{"id": "train_07810", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7cb61e461cc0b6977049e2c011cb4483c82040a"} +{"id": "train_07811", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9750e36bb0a1c48ea6b52e72668fcb95a57e38e3"} +{"id": "train_07812", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "742dc05b5f96b3491fa47c7cb4d7c4a3fa4f01ba"} +{"id": "train_07813", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "863363bb0116dc036835424f8a9f51036b513fdb"} +{"id": "train_07814", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7d2d03d1b3af80158ffb533ff2f784a33d8fe9b"} +{"id": "train_07815", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ff5b70f5f758bb84ca145181e957f99418f9fbb"} +{"id": "train_07816", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fa156ca94506031b6b36b176935cc921beaa006"} +{"id": "train_07817", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "auditability", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26e27d4a81b79c5130998e0c5f4ca2b51fed5d4b"} +{"id": "train_07818", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "425a88cb10f2d4818159eb7349d208dc3d9340bb"} +{"id": "train_07819", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "documentation", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fab301c6d58bcc2446b890448abce64ea2dcb3ef"} +{"id": "train_07820", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eeb35dd2ac08f9873571d2d0caa0527b650e9cba"} +{"id": "train_07821", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbce10eaf65628d0a0a6c6c9317611a80f3b5e45"} +{"id": "train_07822", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc788e17d198436d5092f57410bfb0f44760d279"} +{"id": "train_07823", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2387b0448cc236189b71a208bd426721788553da"} +{"id": "train_07824", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "ci_integration", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d76e7da16e97b55f1fd589fc89459fe39bfa5dc8"} +{"id": "train_07825", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "security_gates", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7819e453f90947fe9df9827e944c0b431e9692ee"} +{"id": "train_07826", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tooling", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da9dbdb92013d2c0dfeec6f2b53404fe6f9a161d"} +{"id": "train_07827", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f9d7c6231418dc32a092a3460a61b6e8eabb66c"} +{"id": "train_07828", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d56427fbf43fe7b079a8ff2227e59c67cebc8857"} +{"id": "train_07829", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0bb12092e10885f1ee29785d25e6b1d1cd06eba"} +{"id": "train_07830", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4628b463ca6a385509cc99acd87649f9401b3063"} +{"id": "train_07831", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b834b6e5eaa9833ebd0358479dd72c4bc515e47a"} +{"id": "train_07832", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "029446c0315b2adae4f68f5c672fbf95d53efe08"} +{"id": "train_07833", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "368bce8a3298bb32ffb36e1110bb7e5f5f63ca0e"} +{"id": "train_07834", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2af3ecc0a2d744ac0b000e8aa1c9ae4d428345ca"} +{"id": "train_07835", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ce7507d7cf30249da9c578d7e81a17353b134bc"} +{"id": "train_07836", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17fba26c3dbf5831ad085aeb1a71ea0555418fdf"} +{"id": "train_07837", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fe14a714e2d923a836e9836a95e9713ca2caa4b"} +{"id": "train_07838", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "security_gates", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fe2e0d1985177a8034fbfde2af340c426650f95"} +{"id": "train_07839", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "documentation", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b270af7ef6e1346eea26689289ec759900e10f60"} +{"id": "train_07840", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b69040b81f77c463b6eed544558f6c805fc04e39"} +{"id": "train_07841", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffb26eae318ebfe5ae53dec789bffe1c057012fd"} +{"id": "train_07842", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6df8d1b76c999a8b089539a525855acf1ff4fa9b"} +{"id": "train_07843", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e696eff0d09b9028b29e9edb218a3a005b915310"} +{"id": "train_07844", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "documentation", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ebeee507d1099954739d0f31b7861c768cbf232"} +{"id": "train_07845", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "273d0aaa92fb4530c4157dd85f662b21c7e0d74e"} +{"id": "train_07846", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "documentation", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da2e27e432c145b661d209768475734294d92470"} +{"id": "train_07847", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "tooling", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d42831b6e9bfd09597a62b1cbca030bd6cdb2614"} +{"id": "train_07848", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "ci_integration", "reproducibility", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3aeb0cbb224e6b8aec630233810f401f5615b1fc"} +{"id": "train_07849", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "ci_integration", "governance", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c179b9a98b8be374ed47365ac3e77b4b29b451ba"} +{"id": "train_07850", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "tooling", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0a444cfc1612802ad5100a8b0e5e5270509be7b"} +{"id": "train_07851", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "auditability", "documentation", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1bcfa1c05071b47f016f0103f100611ca0c99ad"} +{"id": "train_07852", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9fc4fe88ad76094a3c0f0c0be2fff21e23a0eaf"} +{"id": "train_07853", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f35ff22fceed910dfa0cf77e5ca24ffafbb98c5b"} +{"id": "train_07854", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cf3b640d28f27a3d0e1feff2db768e5d9efa6bb"} +{"id": "train_07855", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "ci_integration", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "772014e51e61279602e4156527fe9ce029b23b39"} +{"id": "train_07856", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0289b173820473ad88600be8dc83f4e1bd8e1432"} +{"id": "train_07857", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53b60aa39a6dd4efa24a8cfd945bf015e2f364e7"} +{"id": "train_07858", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f43a03d9395472d6064913f5fd289a849104c4ff"} +{"id": "train_07859", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01118ef55812ff1bc9065e055d94af05fdfe6546"} +{"id": "train_07860", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3218ee57c5faf33dda30d08b0617088ebf55d146"} +{"id": "train_07861", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89429d6948ed68f726df74b29afce98471d1dd5b"} +{"id": "train_07862", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "reproducibility", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfdc02ca850dad0c269e6c96386fef74c3c02404"} +{"id": "train_07863", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26b59b41ced3ebaed653ae2c803e22592deb0da7"} +{"id": "train_07864", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cb13ce8d15420f9eb018b0f4a5c70ade7a8b07d"} +{"id": "train_07865", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "repo_scale_reasoning", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76c675fd66efe4594cdb779b3b07a65e3aca5ee7"} +{"id": "train_07866", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34cccf00d97f1e43d9e44a6f4a9b187f9ad34eb8"} +{"id": "train_07867", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "reproducibility", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "320dd5f7b2c0d674b20972e0f8d98c6994446abb"} +{"id": "train_07868", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9285229c55202fd7e89a2cbf6bb35e847a9521b2"} +{"id": "train_07869", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bce21deebfa1c78f7861f26debaa22b21e6b9fc4"} +{"id": "train_07870", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "security_gates", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d59b4fabaf78dfd076fe37f741c275468390972"} +{"id": "train_07871", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef9f54720b38139fb593f707e8d609f875e39234"} +{"id": "train_07872", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "security_gates", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d84484d139c18b4d1bd82630f9cff9666ffcd9bb"} +{"id": "train_07873", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aab70eba5378d8ea5cf8e87556a44fef351abd8a"} +{"id": "train_07874", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9970e791d5a391a29465d07fdb3b2060a87722a8"} +{"id": "train_07875", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff7683529dc723ce470c05afdd12be5aaf44a885"} +{"id": "train_07876", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "467906217fc7580b801f5a0f1d173779daf9e4df"} +{"id": "train_07877", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "auditability", "tooling", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a757412f5d871cb0ef52e4f6510e1c3d8d521317"} +{"id": "train_07878", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tooling", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d399062816d5d938772e7406cf5c6c47650b61c"} +{"id": "train_07879", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a40cb459224ae5640510babf1701ab8fda9c7841"} +{"id": "train_07880", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "632746ef806ea48679127fb3ef0051e32a3091bc"} +{"id": "train_07881", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "repo_scale_reasoning", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8681fc1d7512a29887208e4c6c6930a65327afe0"} +{"id": "train_07882", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67a4748e8a39ea4796abbcbd8fd0e9f3aa247cc0"} +{"id": "train_07883", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2dcc32d3defbe8d98f118d74360172068ba61b1"} +{"id": "train_07884", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d99f961f9be5b0ac5cf9bc8274210979c50cc569"} +{"id": "train_07885", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "tooling", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d957e31c360ea8552d2d2c44df142ed99744a031"} +{"id": "train_07886", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "ci_integration", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39098818073d0ec670c9c9053b48ae13697766c3"} +{"id": "train_07887", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62936cc5ce4ad14fd71627f11c8a4ee736f6d830"} +{"id": "train_07888", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "security_gates", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2000560a3532c59d9e88e3fd110df792183f5237"} +{"id": "train_07889", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6e4706674a3fe3ac124dc7716c0466dcd71e827"} +{"id": "train_07890", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fea30f83418c8ef97605968f9f84a22a1ae86f8e"} +{"id": "train_07891", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4e7c97cc08109edee7c87a62d1f97c090b8b1e5"} +{"id": "train_07892", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "documentation", "security_gates", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e6c9ade6773c25459568f83122a232d5c211741"} +{"id": "train_07893", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "documentation", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58808287ddd2e84d1b83c1ae64e995f0b85b814f"} +{"id": "train_07894", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd0041ef6398468bb30c4ea05378c599df345fde"} +{"id": "train_07895", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tooling", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "663414870eca80f8f6b6c851993f4543fa066dc4"} +{"id": "train_07896", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "reproducibility", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da40e6d646845b3b9bd36e1b3dc23e567dbbfbe9"} +{"id": "train_07897", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e58b89325927cea63852388633576f8924acc82a"} +{"id": "train_07898", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "571a90bea95767bf0625da74fa1d1f42f42900ec"} +{"id": "train_07899", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9bfd9901a6a907683df316f245af03b282cde432"} +{"id": "train_07900", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4af42fe97d6bbc7d8afb6e37fdda93b098536577"} +{"id": "train_07901", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ac128bd1a493b3b76afbc509dca25a4ab36b449"} +{"id": "train_07902", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "auditability", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a504ee8f25e7a00502418fa9e49a698bfb0a0a86"} +{"id": "train_07903", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed4f8ae03e72439d302ece9a9da0252c7ec58a25"} +{"id": "train_07904", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "auditability", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ad0e3a1b7152bfc9a6633ec4f52d4ca9d4a6e5d"} +{"id": "train_07905", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "security_gates", "ci_integration", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c5e16edd80a760443128b577ebe6b9b221560bc"} +{"id": "train_07906", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "364de191ac5845af0e68385c23a15ed2c14caf92"} +{"id": "train_07907", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff9abd1e2421ab37c7be2d5a8bec3eb8c82d941b"} +{"id": "train_07908", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "ci_integration", "governance", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c4c1beb526a8c6ee36eddb1daffa3e309dbfcf8"} +{"id": "train_07909", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60b221cc2762b433b076374507e6afb2e679c92f"} +{"id": "train_07910", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "governance", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b928f7171c524f2097743553f79f6898b6c68612"} +{"id": "train_07911", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d3225c49b9b847f366bb2022f9bfc3a9c67b51b"} +{"id": "train_07912", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "ci_integration", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47b2628d34b4812f447101602173eb7aaafb5d97"} +{"id": "train_07913", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd954dcb3c9a80932d44a92924ab0b3445e47ea7"} +{"id": "train_07914", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tooling", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "257fa7c05d0d055fdc37b336f214e656e9eb01c9"} +{"id": "train_07915", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea365c375ca654d01bc28d6d58bcca426089005e"} +{"id": "train_07916", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7415773381e5eeb1e3d353f4dbf4f5041dc31454"} +{"id": "train_07917", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "509acf19c89aa304ca560ba8f44ff4663848223d"} +{"id": "train_07918", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "auditability", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d44110c63f93c3387d6af4b66d9237e75771f61e"} +{"id": "train_07919", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad5673c2a037a8fd3b012788a3d668ecc4f5557f"} +{"id": "train_07920", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35e3c373b345a5040c4ca23507271596e144b178"} +{"id": "train_07921", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "895b1e1cc947347695d8e694ba91181d79630187"} +{"id": "train_07922", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6775ea14959ea36586a8c1e411d740f3c995661f"} +{"id": "train_07923", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tests_are_truth", "documentation", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63c2e33f188cf622836a4814ec2f7f055cfa8f42"} +{"id": "train_07924", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07de88a0283d234fafc08c97d840996f2bb3e5ca"} +{"id": "train_07925", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tooling", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8f3a2cacfd25c41668b0ab295d9e0fb56ec38ad"} +{"id": "train_07926", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcfdd02fab16d962137080f1da0e6cc53ccf2ff0"} +{"id": "train_07927", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92499f2746e40fd916cfd355376c9a09c0e55216"} +{"id": "train_07928", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf925f20c09ead6074683f817c212e3abfa5aa20"} +{"id": "train_07929", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a14fc6373cb881c2dd071e1dfc773dff85cad634"} +{"id": "train_07930", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef286bd0132c036b54d269d62535d02b11471943"} +{"id": "train_07931", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "auditability", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b310f11f7f3ef10390e500a8cf7fa903ede3377f"} +{"id": "train_07932", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59a5c7af1adb6959bcc1f57dbef6d4707a2d408c"} +{"id": "train_07933", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e60a641ca130837a8025326e7acafa36bfe2bb4b"} +{"id": "train_07934", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f71a13fb81f1b7ad2b9de16acb822fda2518c16"} +{"id": "train_07935", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "323294c469cc9ca6547d37459fd7133f54fe4ad1"} +{"id": "train_07936", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "tooling", "security_gates", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a350f78821fb775d4994038d511af086f26dabc6"} +{"id": "train_07937", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "auditability", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b0dd3d0c11f4bbe2da50ed323870408a6fe68a6"} +{"id": "train_07938", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f153e2e5db40ee23a4d1d586b68d3d1fbe3fa035"} +{"id": "train_07939", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0194f05c6a51928ad014bacb2b8eeb0d7252e17"} +{"id": "train_07940", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "323149655b55cdf021d874a89584a0e1af28c28b"} +{"id": "train_07941", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b36018df692e45d485fc6fc38ed641eb95c11528"} +{"id": "train_07942", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa83eda8e9b9c3a7029903cb9ea2294b76310c58"} +{"id": "train_07943", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5646dff166dddf8644256cfc7b5775259fc2556"} +{"id": "train_07944", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e176f4614c3e4ea8d3a3cba98f880b4a94aa23c4"} +{"id": "train_07945", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2017d1f1ab3104f75c7c6f1b4a2d65dd8a11a0c"} +{"id": "train_07946", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09a3833b04077d32e3b6abfc6c3fbf1a20dd2749"} +{"id": "train_07947", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "539f6465861c57e3251ab5b95f922b1f63c13cf5"} +{"id": "train_07948", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ac35a60eb4f28a23214356b9dffdc2f15cc88b0"} +{"id": "train_07949", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa5383a5ab8b45bf9d0b355dd8f673da4723bc5b"} +{"id": "train_07950", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14a5aa0903436eddd5fd7ce1904ecc954721054e"} +{"id": "train_07951", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a269d0110434df8cf205bc58ad780d611b345f3"} +{"id": "train_07952", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa4593d51bd89e7502ee192335859267199f239c"} +{"id": "train_07953", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67c6e43020f548ec9e3225f47b69e68678bd1f88"} +{"id": "train_07954", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aad69dfb432afcd08c151f92afd91809824e58ee"} +{"id": "train_07955", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f56bed9b7ce6e49092da83a6570d63af613bd8b8"} +{"id": "train_07956", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "governance", "auditability", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1b62a945fb862d3697f7a3ef931e96739a948f7"} +{"id": "train_07957", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "governance", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48d6e0d5ebf7a2413853b5176d5c5c2b03d0e796"} +{"id": "train_07958", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tooling", "documentation", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c445b07747f4052d15bc3fa4fc7ed14ba62aef4"} +{"id": "train_07959", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a54f769a38a96e9716aa574e895f6fb28ee71d3"} +{"id": "train_07960", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27ea26ba46e5c1fc96fae1489219082ab3062951"} +{"id": "train_07961", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "documentation", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c24a701afef55c87cb58f264ae79202ec0fd28df"} +{"id": "train_07962", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55a869b6853401241f18c95dc0fe0703306da9d5"} +{"id": "train_07963", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "607e76087610319a619ee82d30ce9d313ec19ed1"} +{"id": "train_07964", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "documentation", "governance", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b526a452a286377b2f3401a96adee9e5650981d1"} +{"id": "train_07965", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e48abd6c5fb4bb0c1d7e98efa041c22ac0ec01a8"} +{"id": "train_07966", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "governance", "documentation", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fbb395ec2fbd3c1af41632379d035c9914ab590"} +{"id": "train_07967", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9eda0d2801e1adb768690bf53cc258aa08dca1d2"} +{"id": "train_07968", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "cost_latency_tradeoffs", "tooling", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7028cd1ac431280aaa605e0cbd6d42a78a014dac"} +{"id": "train_07969", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d38fc77de4f8de0dcd600a3afae2901bb9e3c068"} +{"id": "train_07970", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "documentation", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a8e3bdf1876e9dfabd42732701a473bc4db9954"} +{"id": "train_07971", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "324c7cd9afeabab238afe5c71e9fda5172958c18"} +{"id": "train_07972", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "security_gates", "documentation", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0004acf31fb01a12ecd583dd4b4c7343c895d495"} +{"id": "train_07973", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d7dd1021a3535f88be465a5960081ecfc8f5da5"} +{"id": "train_07974", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee3c4e10d06d3e3c60a2c59e6ede1f5f5cc3cbc5"} +{"id": "train_07975", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "tooling", "documentation", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c4760783ecb46e71ff3786e8b60d98c499687dc"} +{"id": "train_07976", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3faf63d0497d1dfb0f0ef1c8739f86d408380a18"} +{"id": "train_07977", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86287ab11aa25c709d2718bfc8b87b9597e8723d"} +{"id": "train_07978", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "governance", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9acac950b622f8ade24a8b56fe52ef55c3a153c6"} +{"id": "train_07979", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "264aa07c9b9b6e5ee6471c3c5bf019a98ddcb0ee"} +{"id": "train_07980", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tooling", "auditability", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c341e76d2d849ffcd9cafd382ac2dfabca66f9e"} +{"id": "train_07981", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9bdb3d5260b3e60dea0619ef08bca3352cb2d6ae"} +{"id": "train_07982", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f0778e8793622f3e565cad2a7caf1770a1fef7d"} +{"id": "train_07983", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "documentation", "governance", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8001ecf168ed6cbe0acf4279c78436087c71753"} +{"id": "train_07984", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1ed93264bf4fdfe138d0c4f879f0f1a363e246b"} +{"id": "train_07985", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "documentation", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4241e4974a9b232c1b444f0282457e7e31d48dd"} +{"id": "train_07986", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f9b775a88bf6ac01394021d7a7f0da09e9e1196"} +{"id": "train_07987", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d459b52ecbf47eb148dbd4196dece79066c81cf"} +{"id": "train_07988", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdb4665545b056d0f661047ed8a57d11a95f1665"} +{"id": "train_07989", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "documentation", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c6856c5871bc8b9f46df5cf33fa1c26a142fcc0"} +{"id": "train_07990", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "documentation", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ec30ef6bffb046f87b97f31ba838c47cc0d0f7c"} +{"id": "train_07991", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "974eb9bb0e474efa52d7ba6bd689913fe242015b"} +{"id": "train_07992", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "auditability", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aeb7905e0ad4703dfb8894aa2d733ab71c562e46"} +{"id": "train_07993", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "auditability", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e91c4d723eb1a21682c6fb975724c83cd9281b2"} +{"id": "train_07994", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "documentation", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e411caaf4983aeb56b67262a165452f93bc3257"} +{"id": "train_07995", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "reproducibility", "documentation", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43bc658c165b79b3738c81fa5a77dbe4cadb0119"} +{"id": "train_07996", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7594fa8fcd18fbef21c4e6da5e9e9822a98c79e6"} +{"id": "train_07997", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7c15bbda525a7fcbadb92270d7735aaceb45e09"} +{"id": "train_07998", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb8a37fd91d744a7bb38f7cf834855d58fed552d"} +{"id": "train_07999", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da45109f17322c70695b6f75e3432bc5cf18b2c8"} +{"id": "train_08000", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "documentation", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1511f4fbdfefdc7a1f0a7c90692bd3679deb01e3"} +{"id": "train_08001", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "repo_scale_reasoning", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a83bfe25b3f75700434adcf3a0ab7575fdd537d3"} +{"id": "train_08002", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1485357a47f9da9ab843d2ddbd073c82904df452"} +{"id": "train_08003", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "079fdc8fe3c4f403ee1ed1a1284fbe27d45cd64e"} +{"id": "train_08004", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0a85391fef5c8e3f8c2cbf6921151e6ad966a6e"} +{"id": "train_08005", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d90845ef073529a9ea24da0328ecc8de051ea498"} +{"id": "train_08006", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "ci_integration", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d818f2a608ceb8cb599a761b803f4a516e265c20"} +{"id": "train_08007", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c709e9e03ef6c268561c2e0d5305fb60cc5d853"} +{"id": "train_08008", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "governance", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eafe37a40e7dcdc45ed76e98be4d8b74401bd9b9"} +{"id": "train_08009", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ae01491838a155f91980c42a857af540d9e840b"} +{"id": "train_08010", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa5baf58625c496c674722c1bb615300403ed14e"} +{"id": "train_08011", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5eccef89866d1aa7653abcb076630d95a799737"} +{"id": "train_08012", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68c0e895e704c5915e5e7c6aed39f26df7c879e4"} +{"id": "train_08013", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "documentation", "governance", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b5670a05efc38ad73d6c660c4c1cb3818b9c684"} +{"id": "train_08014", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f4d26d1eb57c15c768cb3761e8565b97d0293e9"} +{"id": "train_08015", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5ebd2d6465066623a32f7e1074b5d6171120fc7"} +{"id": "train_08016", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "465f56f983c325a357db8c82121bdb2843382352"} +{"id": "train_08017", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aaad428bd4f9ac772a607243114c7377a2eac8ab"} +{"id": "train_08018", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9583daf7bccaf0554a2dcc2310fb7f35812efda9"} +{"id": "train_08019", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bccf9052f4fd0a5bc0127ac0225e2f0f71e6fd78"} +{"id": "train_08020", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "security_gates", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe887e6cae86df26ed64d90fe33358860f870b5d"} +{"id": "train_08021", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d559bf3f91286992d55d6679105f32afcdbd13b4"} +{"id": "train_08022", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04b4e73c335db953c4c160d380ae3ae7b3a47397"} +{"id": "train_08023", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "governance", "reproducibility", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdd740fabd2bbcd410805752011e315e378e02ae"} +{"id": "train_08024", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "governance", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2852262ff75162328b4324c962d12c01ae64216"} +{"id": "train_08025", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e248886462738d13c07f282c470da9389e8973ee"} +{"id": "train_08026", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf2a236fe311e0471683d3572f1a01da554c1db7"} +{"id": "train_08027", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "ci_integration", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4486d18e6867856c1b903978753f2984d8343761"} +{"id": "train_08028", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74d2b0fd1e15b3ad4bacfe8e73f889cf0d82ca4d"} +{"id": "train_08029", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96d4fdff6527549f2d823c7998186a1a3fbe11d0"} +{"id": "train_08030", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d8ef1ee7e2035551735d6de8238aabf3b4c4fe3"} +{"id": "train_08031", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5810b4baf86072dd48807c7c97d87b3c451bc5fa"} +{"id": "train_08032", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a614f00861ff5b29ca96458e719b7a59f43d6bc3"} +{"id": "train_08033", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53d6f61e28f241b397d2a55bab8aa6051dcb2760"} +{"id": "train_08034", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4be3712f9995664f320d2699227e29d9fe2bd99b"} +{"id": "train_08035", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88e20ccf243b3491d14b1c44d32c068d81effcf7"} +{"id": "train_08036", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f30b85de6066fbda12c6b1adfc6c3e64dece3121"} +{"id": "train_08037", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bbeead378deeb68bc7282336fff0d0b011907fe"} +{"id": "train_08038", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ad6950d291469fa61453a1db678e46fc8d068c4"} +{"id": "train_08039", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10a38a1d5d1ec2c5a39712699de9a5fa5c8e7b94"} +{"id": "train_08040", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "438f37efadcc9c33da7a8d58f2da8841d840acef"} +{"id": "train_08041", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "496d95f76d8a4412af0d1f426e8a7b3d08c7e722"} +{"id": "train_08042", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a3a96c5ea68c0e16e19da1263e2165ec5e0265a"} +{"id": "train_08043", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b60a149558ae3ad18a1708b9caee9794470b088"} +{"id": "train_08044", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5840a595c29379955af6d79fcd45941d7d67fb3"} +{"id": "train_08045", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de3220e90f0bd956dfcc6d80cbc46b16ea9fad4f"} +{"id": "train_08046", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "tooling", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed07d2af8b41133b143c0c53601b417bf3d54d96"} +{"id": "train_08047", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "072ceccf970e7af468a6fbafefa9a50665eceaf3"} +{"id": "train_08048", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c6e4dc7059f84c3105265d5fe7b8091cc6213a4"} +{"id": "train_08049", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "tooling", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca5e5447cbbdf790849d38141c4ff59f07e5756c"} +{"id": "train_08050", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "documentation", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e498ae8b098874941974cff60ae1b3fff94bd02e"} +{"id": "train_08051", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "governance", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b97edd173f1589b226cbaf1e231c364947739ad3"} +{"id": "train_08052", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "tooling", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85b0ac335f1860b13b49ac2cf58b388d45e73bf6"} +{"id": "train_08053", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b644826a8eb827f86350783e8826beb38763e06f"} +{"id": "train_08054", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "evaluation_metrics", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f6359dfd71813231a1369c61f0af31b2b9aa991"} +{"id": "train_08055", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d26a917da257fbec4600bc9d2ff775b78de2ca38"} +{"id": "train_08056", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64b8049780989337f15324b95064371b31529467"} +{"id": "train_08057", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67b4839cd5c6e744a9d3dd02668e36dff96955f4"} +{"id": "train_08058", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1fedecfc48d2b7f3266d98394ea1ad4f3476576"} +{"id": "train_08059", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85d824f3453c08256a0b4a371cf51eb0cea37a5a"} +{"id": "train_08060", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a8a669a04d371301f7e7eaee6f0ae3d06e020c7"} +{"id": "train_08061", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3932d7d06dfda1e1df51adb0db880f65cf0f919"} +{"id": "train_08062", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c6e27c5ba7d031e29c9a7cd0b4ec064dc5f0d2a"} +{"id": "train_08063", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b243227be1d35665108498112b1dd023849ca01b"} +{"id": "train_08064", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "evaluation_metrics", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffd06b0f95d0a277999858e8eea60afe2dc30c0b"} +{"id": "train_08065", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "ci_integration", "auditability", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aab25c26ad9f6b8e2f9b7b56c460fffc5c584d4e"} +{"id": "train_08066", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb1c319751c2872da4bfc5782b732f179ee9410b"} +{"id": "train_08067", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "249e8178108b55423573a61c2f61a48baef62506"} +{"id": "train_08068", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "governance", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b0dd3d0c11f4bbe2da50ed323870408a6fe68a6"} +{"id": "train_08069", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "governance", "documentation", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58f922ea7e8ab74835175e5687836082910ad410"} +{"id": "train_08070", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34999d636984facf545aae4785582daecb3790fe"} +{"id": "train_08071", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daeeee609f8af1fd6dc916095e3a6220ee377189"} +{"id": "train_08072", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f511949ed9d49cd0c666f1bab4a2d683a8df7ed"} +{"id": "train_08073", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "auditability", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb1ded83e66eacd2de38957cd00f2c1fec0b2d2c"} +{"id": "train_08074", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25e0ded74f141cd6cc2d8295766d261ba25aeae7"} +{"id": "train_08075", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6a14215fdb4f0ac5e8cc253acbd608cdf1bb2f5"} +{"id": "train_08076", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8bf444c6afa706b79f2ebb7f683394fa4ed6bf9"} +{"id": "train_08077", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c52aaf9b1dc38c64eca808502da6cebdb23933a6"} +{"id": "train_08078", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "reproducibility", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42cdc6dba5b0c0ec210fca875c6110f5cb85c246"} +{"id": "train_08079", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ac837a1554111a9c6fd427c142fa90f41424f6b"} +{"id": "train_08080", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "tooling", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db958eba98718c8f67ae822df03851b974d0e40d"} +{"id": "train_08081", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cbfe77fd04e38108d7d445c4a53d0554f6d8182"} +{"id": "train_08082", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e9300ef846366a9963e85dbeed23c2d33d40899"} +{"id": "train_08083", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1eb1d563327fc3e49052e9863589adf493184346"} +{"id": "train_08084", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce380ba79878705d8b9fa8e9a0a6f9ada25fcf27"} +{"id": "train_08085", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "auditability", "reproducibility", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6bcaf935e634dd85a2bb63eb8162e36abb30533"} +{"id": "train_08086", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41b633f26b9029c844af300e1153ca914d6c6869"} +{"id": "train_08087", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "735db2e5f6c2c3163491b5bc41dc3b7e0b91e178"} +{"id": "train_08088", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3408143855076a1a9949e121b5a874c415e4d36"} +{"id": "train_08089", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34599deb72eda0fb5c23b2bdafe03da3a39bdf57"} +{"id": "train_08090", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c45397c1adcab884131d022ff1e346712a12c1e9"} +{"id": "train_08091", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "documentation", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d991c6ea4bcd056bc06335b7264f4f87465d1e6"} +{"id": "train_08092", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d52fce3ec1f4ed379c8bbdeee3459a39ccb4392b"} +{"id": "train_08093", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52817542c23032420f74675680881574bf5eed9d"} +{"id": "train_08094", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1436857db7dcbf4c855f60dce6f247f1bf9f1f0c"} +{"id": "train_08095", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tooling", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7060635b2621a39379fa9c889bcb35cc846e3d14"} +{"id": "train_08096", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eabf523762507f8b36a984c718070334693d41d1"} +{"id": "train_08097", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2614f6e3e42a1653399729213f2db826a9161d3"} +{"id": "train_08098", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60a9610921d20f783e48d8433cad2e9441625062"} +{"id": "train_08099", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6561143872d901ff4ef3deb04d4d02b83bb64895"} +{"id": "train_08100", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf79aa361c239ab94768f67979d9a6c92d1f2a92"} +{"id": "train_08101", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "001df8d625420b32aaba69ce9b2248774ebe543d"} +{"id": "train_08102", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fc5be51de9b6a82a541a54d4659772d4ebd7966"} +{"id": "train_08103", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fea0bd741622a3dff2cb38856dc1b49232f37001"} +{"id": "train_08104", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "auditability", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43604c89a72bf09fd3e6e4a345cb09fd90bf97c3"} +{"id": "train_08105", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d48c3c07f2fac782cbeb758d49f4818dae28c19a"} +{"id": "train_08106", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "reproducibility", "auditability", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27ec5bceaac023c5736839ad753e79055a04c8d5"} +{"id": "train_08107", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c45071c31d39f7bec084a48276b42fa695636800"} +{"id": "train_08108", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "609fb123da95bb02881c87d4f3fa32fc7373a3db"} +{"id": "train_08109", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9882e30b93e491e36da037a008e398225708e936"} +{"id": "train_08110", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "ci_integration", "auditability", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef1263083145c2daca406b13fb9e4609f259aeb"} +{"id": "train_08111", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26e602871192c7239f6a8d7af9682d63b9ffc038"} +{"id": "train_08112", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5e6d5703ba624a947bebd26137bb4bc38c06798"} +{"id": "train_08113", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a209398280ddc15f8cd997f45911fbd1eee3d03"} +{"id": "train_08114", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec74a038d41f1a938e1453ee1f7157a395c052ed"} +{"id": "train_08115", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34542abf6919fcd14f45b88064120bd515b43750"} +{"id": "train_08116", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "ci_integration", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a4b7726d444df71f6f57726f46da4d4c2d92211"} +{"id": "train_08117", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "repo_scale_reasoning", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18fc4974b7bdfc23e76344dba5f5b034808de97e"} +{"id": "train_08118", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34fb5c6f47c5258cd9ba2bfa4b12559c7d6a2241"} +{"id": "train_08119", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a970e16c5d4cbe7655a419b8c5968bc1b15e921"} +{"id": "train_08120", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c257eca5d519042d0b3e83633f442edc47c43a03"} +{"id": "train_08121", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "security_gates", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90f714ddca20ddf275703e3073d223437e1e62f0"} +{"id": "train_08122", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07a738b69b6a0260481cf61c8c46120fdc63eff1"} +{"id": "train_08123", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcee2f409367b7cbe9a7211981e527d639b797fb"} +{"id": "train_08124", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db1ebd501ab9892dd1c98785de2f30629350312c"} +{"id": "train_08125", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca4f2386e34b94c8d394c000f9b2afa8d2316bb3"} +{"id": "train_08126", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "governance", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63540623b377d55e6c9a505b768f486971fa089a"} +{"id": "train_08127", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2510273386d3de29d285b04256042203b4a4075e"} +{"id": "train_08128", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c92b64139db7ebd648739e1b7976a753b732b864"} +{"id": "train_08129", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "tests_are_truth", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0339c84f5835afa3decf1f286d5387c1d1ec7779"} +{"id": "train_08130", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cb41868de41a4990a125354269d89cd3032edce"} +{"id": "train_08131", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfe7aff556eadb3f5a3091f6fc4dbd78e210528c"} +{"id": "train_08132", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc628b513931eb68b8d4a51b66c6ddf592b54574"} +{"id": "train_08133", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "governance", "evaluation_metrics", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be6b244bfa0442b300b346b69e1c25ed2647d1b2"} +{"id": "train_08134", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53bfaca298ac755cafce0b158551a6ab8c41c191"} +{"id": "train_08135", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f33d4d7d3c0d178787dc88a3e0646650c76c5199"} +{"id": "train_08136", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "security_gates", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04b3de7c43aa03f35044358a12180d9e4608fa45"} +{"id": "train_08137", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae3aecb9aab3a4be501ca0198cc1824f2459bc11"} +{"id": "train_08138", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4fd83a756d23be0ea909dd7efe40f3af4abaa8f"} +{"id": "train_08139", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ef28da35cd2a5c987e6612b81e541f39875bfb5"} +{"id": "train_08140", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "documentation", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9de92d6a2dbcd4b2d2ac3e07b7d48057c4f5b3c3"} +{"id": "train_08141", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bb514b0807fb752348345d52435f42d891a3ffe"} +{"id": "train_08142", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "documentation", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42184457a92ecfe8e4f9635fb42650906215b721"} +{"id": "train_08143", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "documentation", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a47dfb1109831e4c40d3aa436ef72fdc60b48026"} +{"id": "train_08144", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80c6862b0dee3fa50473a19671fd932a7dc388c3"} +{"id": "train_08145", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "reproducibility", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a658dd9c710b39d070eaffa26ce9756b0bb5e97"} +{"id": "train_08146", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1bdb886286cf3a923b23afd0dbe7bfa2f8368941"} +{"id": "train_08147", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4bc059eee9c9eb49d5bde36f2718b950a0a7fe14"} +{"id": "train_08148", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72357dd750fbb90ce1ecbba802a1cad1665a2dfa"} +{"id": "train_08149", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b3b980c7b1ac6465a67ce8ba3e9b04254ff5a08"} +{"id": "train_08150", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72d128472188ff21e6f286c6dce5817357b3637e"} +{"id": "train_08151", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09f4a8a9ce80328cc420a4235e429a55e11c66e9"} +{"id": "train_08152", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "security_gates", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2195bce3a1ebe305babcff3424867d62d6a3aa86"} +{"id": "train_08153", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "security_gates", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86c7221bd7dafb147be9d2e7adf0a1c723d0db17"} +{"id": "train_08154", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "governance", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "453c6647ae67b106a693a1a022911e6b2bd0ad9e"} +{"id": "train_08155", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3583e387a946a1c58b76b5d299915723546abff8"} +{"id": "train_08156", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58e0b81ec58196536575947ef1fad5ff8efc47c0"} +{"id": "train_08157", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tooling", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37c0feb37cbf5d9d9e7fe862d55ef7dbb4a5e17e"} +{"id": "train_08158", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "ci_integration", "documentation", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "843bcc86f6752dd1a8aa48cca36ebbeedab5c1c6"} +{"id": "train_08159", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d6f669eb9ae148417aa7f452c7d0c3d0878ed21"} +{"id": "train_08160", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "reproducibility", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c9fc0bf9549ce1473b034bc1ec6eec728edb2a5"} +{"id": "train_08161", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "703729f8222cb1ea6d77722356dddbea76053515"} +{"id": "train_08162", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aeca78ea674df7e29e6ba39cd9c2d548c2ff31b4"} +{"id": "train_08163", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "522dd8179c510e5d4519301d04228eab536f5c55"} +{"id": "train_08164", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "cost_latency_tradeoffs", "documentation", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "898ed17f5ab4fd001ab590b4715725df796bb185"} +{"id": "train_08165", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89fdce024dcc0299dc05a03a0d25247dc93c183c"} +{"id": "train_08166", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "reproducibility", "ci_integration", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d6e4473516dc4d3afc2d247a12bdf4763e188d9"} +{"id": "train_08167", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tests_are_truth", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f381f43d40fda4dbe816db7cb78a7707952ff585"} +{"id": "train_08168", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4f580f008b73448c9e416c0a042ee4699b7c5cb"} +{"id": "train_08169", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc72c192bae800a8215212cc82b57ea7c09ea2a2"} +{"id": "train_08170", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "documentation", "tooling", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdc3032216d4a5ba022260ecd5075a3621586ec7"} +{"id": "train_08171", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b526a452a286377b2f3401a96adee9e5650981d1"} +{"id": "train_08172", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c15345d8f2bc534a3815bd033b1e5eb968cf6af"} +{"id": "train_08173", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd519b1dbf4165d34bedabad58fead50557ca578"} +{"id": "train_08174", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53b256c53d05e666131c3b93e1fb857a05136598"} +{"id": "train_08175", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "ci_integration", "governance", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b47f260bc969da5a85dfefff464e4bc35102af12"} +{"id": "train_08176", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "reproducibility", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f938c6c13a03e29d8aafb48143aa2f1c0407e889"} +{"id": "train_08177", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "tooling", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b174c4706564962926a819c5f8e3df2cc1a3c64f"} +{"id": "train_08178", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d7ef00736e51115c757136fae3b01bfaba70567"} +{"id": "train_08179", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "evaluation_metrics", "ci_integration", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86f4d06124b09c086c1abceddf458e544f2daed0"} +{"id": "train_08180", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d8816e55b59eea212fc76b148a63e11d73795b0"} +{"id": "train_08181", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tooling", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7d056e11c1c09ce2e6c2e445aca2fe6705484ed"} +{"id": "train_08182", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "security_gates", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a154dc19b71ec4afec3399468d9f402bd81fa352"} +{"id": "train_08183", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdde8b5c36cb59ae1f692edadb3dd9a16b977780"} +{"id": "train_08184", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5cc79a81ccd9574a32e8438baf596c3e2a900175"} +{"id": "train_08185", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "governance", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1c2040af1f2e8b03d313ec64c4be68d60bc2056"} +{"id": "train_08186", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24a29052d16b2971b10a19c139d248ad8101035b"} +{"id": "train_08187", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "ci_integration", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd2e0671324624cd71d61ff66e905274d74c56c5"} +{"id": "train_08188", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a06244f23fc693d1f724f1aecb8136f2e29b7f33"} +{"id": "train_08189", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "tooling", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4e6cd030c360ecf99d9da01cbd11033e5c4c3ec"} +{"id": "train_08190", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tooling", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c86c638e5854cce751370e0c77b6b5c89b15746"} +{"id": "train_08191", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "documentation", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d198ab8c1c9fe6199a855de8bebce27ee402bf7"} +{"id": "train_08192", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d156c961662d48f6a35f22989a869cde17b38fb"} +{"id": "train_08193", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "governance", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "beb41ec901300b8206653ab896d993c032b60c1d"} +{"id": "train_08194", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "governance", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c9d7bc700686d378e44c10904c6a23707a0f082"} +{"id": "train_08195", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ba9a1d300bed3758be39b89b9b2312c343b516e"} +{"id": "train_08196", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59ee8b87d26e6796072404a0625d87a0aae38c7c"} +{"id": "train_08197", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "027e19f9ac59c16bde408891c73074ca94f442c1"} +{"id": "train_08198", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e17df62d5d3560de8d02576d4d8cf880d589c5a1"} +{"id": "train_08199", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8082a04774046bc742fde287d80210bcd212be29"} +{"id": "train_08200", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31b70315c7753574af5243747a0196d0182d4470"} +{"id": "train_08201", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e709e9e4a3830e7da345448da96b4d64dc539a6b"} +{"id": "train_08202", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "abb3da8d765df5f6b947c70baef2d7bdae47f205"} +{"id": "train_08203", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "adf12eade4082f92ca4e81a06f6402c73487ee3b"} +{"id": "train_08204", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "tooling", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba24c042848c2a8bf7a0aed7b71904283e22a2ca"} +{"id": "train_08205", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0a919d8133801aa3234651230d2630a8e4029f9"} +{"id": "train_08206", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "governance", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f287d7518c6924ef9c2cb81242dfb03d99518b66"} +{"id": "train_08207", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ca410e5c81e7ae7912d15d01d838e0ee541e0a0"} +{"id": "train_08208", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9d0df2155b693604e1571b2290576c8ecd73727"} +{"id": "train_08209", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "security_gates", "auditability", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1c31949bd9519f5d8c46f7ed63a1977fbe3e1eb"} +{"id": "train_08210", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee72424210b9af3abc4f751fa0ff5fd9d201e127"} +{"id": "train_08211", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7bdabda7066ea2230e84b39d926b7337c1a7043"} +{"id": "train_08212", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "reproducibility", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01c6a988f94d03b97450b4059c38940ecd5872a3"} +{"id": "train_08213", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "173fd23a9efe25d66a612ee863f2fc613f2cf10c"} +{"id": "train_08214", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78de76474794f5cbfc0607923d185f6e6b6ca125"} +{"id": "train_08215", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3cf729842a38a86794223d3272496dcbd3df640"} +{"id": "train_08216", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30b8178dae4d47af16670bfb7a1ebb3174388f83"} +{"id": "train_08217", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9eb4fd6514873538b5f1ca8ea4005096a20c1018"} +{"id": "train_08218", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6190ca8f8f5329b1d0e168bfb6dc3e04fa547661"} +{"id": "train_08219", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59dda8f1fd0472f32d82d7e60644949f99c2cb06"} +{"id": "train_08220", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "reproducibility", "documentation", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49299997440697d73b2de33e8acd6788db0469a2"} +{"id": "train_08221", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "tooling", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffe8ea4fb03d500fc5ba1508b6166a0c0e8a0450"} +{"id": "train_08222", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62f0bcad36891d67012cea4a580bb2f157419ca0"} +{"id": "train_08223", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "governance", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60099fab64bbeb6c1f966eb25b79d8dff5248479"} +{"id": "train_08224", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c49abc24076dce99d757a3b6b7f4d8b1fcd766c"} +{"id": "train_08225", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tooling", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fedd81a6a619330a685d9dd52e1a9be9b74bffa0"} +{"id": "train_08226", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0048108b6d503b677095fffbd467bf151e69fb82"} +{"id": "train_08227", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab9c3c5027ac1e027dc8ec3cc2bfd3022a0ac6a7"} +{"id": "train_08228", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdf2d661b99fee070240ddcfb81d8609a1fae1e9"} +{"id": "train_08229", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b7ab4db48b6bf91e1aafdaf8bee2ec70dfafc36"} +{"id": "train_08230", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c5ca8edd44b03c64491d722a85df4923f5cf7bf"} +{"id": "train_08231", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "auditability", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72e0a15cb0d2bd4c60cb8f3d733e59b69f407634"} +{"id": "train_08232", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec4239d7dd33a08460bae772fa6e8fdd50d1125e"} +{"id": "train_08233", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "governance", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f85d27d266b19c011a37f018c7549dddb0a9bed5"} +{"id": "train_08234", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "edf5a108d7e1e0158248e546639e154d7f5d55b3"} +{"id": "train_08235", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4d5c38e7f7ecb7b9131416143da8bb3911a3eb7"} +{"id": "train_08236", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77984ad8f969f28417f4b0a75da3d02689396a19"} +{"id": "train_08237", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a86da61187b5c0caab8701754a357ca827a782cc"} +{"id": "train_08238", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d051487c6dab80d3b0635e901658ff535c989d56"} +{"id": "train_08239", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f788154e02769b299a22677fdabe49d26c3029a"} +{"id": "train_08240", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a0614a25ed381745bf60de73d0d2e771aa76f9c"} +{"id": "train_08241", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "governance", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ca7db75dadf8dd896bd81f126212c8ca8f81d55"} +{"id": "train_08242", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a7527abe57685af57cf511e8f2803074a9c19a2"} +{"id": "train_08243", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90b6c8c2c501611aff7e627829ad7a52ab2e2974"} +{"id": "train_08244", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "770e5eff2e67ff12e07b946787b1bc0b92e0128c"} +{"id": "train_08245", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7feb2602912d2c5700d33725029643b0bbf97dff"} +{"id": "train_08246", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "reproducibility", "auditability", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b33c215031445e791a4de43f8f85fd3988dcfd3d"} +{"id": "train_08247", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79cbb50b146db0dd4c76f675201087b01c45a38e"} +{"id": "train_08248", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cbfee780d03e5602e3a9410ded5da4806ecf88b"} +{"id": "train_08249", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdad78149e4ebff93301b9a891b9d14bc85e623e"} +{"id": "train_08250", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "ci_integration", "auditability", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83bac1bb11180a67e964acb2adad99034e217923"} +{"id": "train_08251", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tooling", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c9ecfdccd5c196fdfbf331621ef668a13db2a5f"} +{"id": "train_08252", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "governance", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28066dacc105e862d47b11c9b01b767cdd31413f"} +{"id": "train_08253", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e409b9d61cbf8a3d9eb163be6f1e4e8a466c593e"} +{"id": "train_08254", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f19c5041d1de88692c48cd32ca50e74a2a9ef3dd"} +{"id": "train_08255", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99b7b3800dc125a24685e363ff29987866aba6eb"} +{"id": "train_08256", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "tooling", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "041d9f4ec29572e4e26c59ee12cc35d3cd022afd"} +{"id": "train_08257", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "documentation", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42d7ff0639d692034bb529206a9f28b2dc7ce5da"} +{"id": "train_08258", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "governance", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a258093e2c3c4cceabf786b4164ee6a4a02396b2"} +{"id": "train_08259", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15c8edc47de1af67786ebc80e5896d3bfa833d6b"} +{"id": "train_08260", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "461e2d5858dfde19b736653b36abf8bb7d7d335b"} +{"id": "train_08261", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9565440dbd6c245750596a137a40ebde11e6a617"} +{"id": "train_08262", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "evaluation_metrics", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c355739a274402fa08cc2c4fe60446f00c3404d5"} +{"id": "train_08263", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13a9766e775915241f7d4cf404778c381e9f0840"} +{"id": "train_08264", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "documentation", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f926d59708b3944153b23db9b9e1fba5fdd80552"} +{"id": "train_08265", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7a286831ae42f8b1f82a56eb50e733b542939e0"} +{"id": "train_08266", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f26ab9416b7d5cdc28cab99771262e8f293469ec"} +{"id": "train_08267", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "ci_integration", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b83f8401a8a702ec016bc4c5eea2d2b852959ff"} +{"id": "train_08268", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3cf729842a38a86794223d3272496dcbd3df640"} +{"id": "train_08269", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fda755bd8a14cdbc4730c7e1c533bec6fd90a476"} +{"id": "train_08270", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "evaluation_metrics", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdfbf2bb23f99fa0c009ff06441a9b6077b1dff8"} +{"id": "train_08271", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "tests_are_truth", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6d8a217f2fda8bb0a9cb1b3b1a134a9f03fba31"} +{"id": "train_08272", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "ci_integration", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfa1176d12372a9da1a496f934d2c0d327545410"} +{"id": "train_08273", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70358dbd5cb00f3f6e6cc90c86485d6d42c6db6c"} +{"id": "train_08274", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb1a98b7ca24f295fed3236dadf4561b297b60e0"} +{"id": "train_08275", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b6b58981169ceda19e7fd49777343862d4df4c8"} +{"id": "train_08276", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "auditability", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2eb2aed57df9c6c5cb588142bca60ac1aee2f53"} +{"id": "train_08277", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "auditability", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e79d666774a18a54fb4854ab3d661ef15a4eabc0"} +{"id": "train_08278", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tooling", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "effda1f0c3dd0225e7aabeed8e6a5b2dbd68da5b"} +{"id": "train_08279", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "reproducibility", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "363aded3bbef925fbf26645b535fda72d777e960"} +{"id": "train_08280", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "governance", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0d1500829c49eabf8236be6ec005cfadee979a3"} +{"id": "train_08281", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c01d767bf25bca92e2e0a00639fcee27df00bf5"} +{"id": "train_08282", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tests_are_truth", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c39d55761ff3b53a7603c625c07ace8a914c9077"} +{"id": "train_08283", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4eb3f82a9d825e7e1957c704a44c2116d1a64bc"} +{"id": "train_08284", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4990b4fc17d5f182b30d3ae8b308f9d54304be1d"} +{"id": "train_08285", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "32a7a6588521854009aab85f007a17c3531d8368"} +{"id": "train_08286", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a37899ddb5cf4ab6753e3f3a825bac14ba385c16"} +{"id": "train_08287", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b8041217ade7715ec348fe5854ee0788ba2c8ad"} +{"id": "train_08288", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6614c1fbe3196ee6e8719619be3b71919cac58b"} +{"id": "train_08289", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "988b6f058f6f5b142fc11750275c4b08e25ed6e7"} +{"id": "train_08290", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d7a04e92b446d6a1d3d350e3af137f51f6b7898"} +{"id": "train_08291", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7467f2b9083df814e317193d54aab2ea3f28949e"} +{"id": "train_08292", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcdd1cf58908279001c7214be5c47352dc58b0ff"} +{"id": "train_08293", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90f714ddca20ddf275703e3073d223437e1e62f0"} +{"id": "train_08294", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "documentation", "security_gates", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa8d4782003955cf5f8042fddcce272553473bb1"} +{"id": "train_08295", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1e4459347b20020552aaa7b57d836300b6e6c4c"} +{"id": "train_08296", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "tooling", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e66ab97d6ce31b7464d932aa189f1c980bb3c59d"} +{"id": "train_08297", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0789c2d25fa9089e95e42b3e62a37e83ca3801e"} +{"id": "train_08298", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8af87d1e408fd5cb486a26d224f913488ee49a3e"} +{"id": "train_08299", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "reproducibility", "governance", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4e1deb77a39322050ed9d617cf3995427c4fa42"} +{"id": "train_08300", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e57d2f2b459c32d7f2994a3c12021c7421490a2"} +{"id": "train_08301", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb10fb0afc24df06b713e6ffca90120bcee30440"} +{"id": "train_08302", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3646a8d8f837be0738e428e7a72c4eaac62dedfb"} +{"id": "train_08303", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9b7361fcc9cb465a2eedaa66ab49a47836248c5"} +{"id": "train_08304", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95dd76bbea7805d47a0ef9f219913cf27c7321ab"} +{"id": "train_08305", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "798289d23be065350d6fd9152bd6880c46154a54"} +{"id": "train_08306", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be392eba4a20b071198ead54fb3d0bfa6c90df05"} +{"id": "train_08307", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e340ea86ca06d6a0d743a0bc116ff2c7736fdc4f"} +{"id": "train_08308", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b99a434a4084d394e5fec760d332f411eb98d5a4"} +{"id": "train_08309", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44f5e80d0db7435c849aed0e151e5d07f186d92c"} +{"id": "train_08310", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tests_are_truth", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6773c99db1de9ebcea8639e2c8f1da50047b6ac"} +{"id": "train_08311", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e842911fdc08e4521fcbbecfb3fbde820b9750e4"} +{"id": "train_08312", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d72900020e716aa3daccf891f8ef0f924f9905b"} +{"id": "train_08313", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "repo_scale_reasoning", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3fdfb4642ada79537a8ef5a536adf9e16af61c1"} +{"id": "train_08314", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fa89533996246f7e6aad57117025622065e111b"} +{"id": "train_08315", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "ci_integration", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fbfe7603e86439b1c738f79917948ac85100460"} +{"id": "train_08316", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "ci_integration", "tooling", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85e11ad646f178713dd7efa0a0ce46293c0a5ab8"} +{"id": "train_08317", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a79641c6fae25c5049802c89c4aeed94e8655341"} +{"id": "train_08318", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "auditability", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa545fff4caa8a38ced4b1d09b856157a4f68de3"} +{"id": "train_08319", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "auditability", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af1ca0ecfbe1b572689624baa59aebaf78a1cd31"} +{"id": "train_08320", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dc142c0eee69afa68bfa329af00d65d45af8a30"} +{"id": "train_08321", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "ci_integration", "auditability", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06133f698008984cbb193df66256a0e434ec014d"} +{"id": "train_08322", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb122212f2311446ff0a59e4e42d23c039d496b3"} +{"id": "train_08323", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tooling", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "694609446e654e645d5e458752231ebb599c588c"} +{"id": "train_08324", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86e9d8b30cb69e9a892f8ec90b3f9c2faa700241"} +{"id": "train_08325", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "auditability", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b568695557495741a4a6aa5dda9d3bbb4ff52e9a"} +{"id": "train_08326", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ad0fd1c003cd10056eb669ea6e2c7fffe560a50"} +{"id": "train_08327", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68085678c5dc3be9b15d5276a28963e3c79f67c2"} +{"id": "train_08328", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a326dc600e18cff9b68f8f7f173ff41b27adbfc"} +{"id": "train_08329", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "149e173e194bd55408d1444ce9690b9c96e1a483"} +{"id": "train_08330", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53eabb4dc28ecbb3b9f8eb53ccf891ecd74b2f69"} +{"id": "train_08331", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "documentation", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcf8a20f9ce11900c02f2fd606cf85fe0ae6eaab"} +{"id": "train_08332", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "962c12729106230186956fdac8d1b1251525ac28"} +{"id": "train_08333", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86e8feb58091ad990a926affed7fa3e9e6d3a5ff"} +{"id": "train_08334", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "security_gates", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea481c98dd4f98a52cfcf20886cf4c64d8842bcb"} +{"id": "train_08335", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "documentation", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e6ad4179389366c2421a49e301576292ba5650c"} +{"id": "train_08336", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91b22f68ea9214948e39b262402a5fd4cb89458b"} +{"id": "train_08337", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b335e901c2e5d0885d12d50ab5854f61b2202bdb"} +{"id": "train_08338", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b03b213e601b98319255bff259090fc268c25e73"} +{"id": "train_08339", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tooling", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a82fd8995d8a78d163fb227ded675001828931cf"} +{"id": "train_08340", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bda80f03d010bf1a3575fa3e26f62c4ecc07f264"} +{"id": "train_08341", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8943c336f8efefaa9fe9ce4399428a943d99bd14"} +{"id": "train_08342", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e12753f50778884d03bd4d14948e0784ffc484b"} +{"id": "train_08343", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cffca0f7e75c59824856f4aa2725a5df41ad8ba"} +{"id": "train_08344", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "security_gates", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3543ea9e9b28cace873220a9d326c35e7d8711f2"} +{"id": "train_08345", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tooling", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "611c681549d82a0f0fe822c856f32c2bfd599f80"} +{"id": "train_08346", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50aac3723e8088e5a0d5265eb0489ff0084aac20"} +{"id": "train_08347", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3ade5d02440608b7d13c4704eb9d12beae196b8"} +{"id": "train_08348", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1304b299c2d772661ce688ce3920848614581d7e"} +{"id": "train_08349", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce3da8c12e160fbd979ed63925ca94a0057ec627"} +{"id": "train_08350", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e951ac0833b2df5d32334ecb630b561dc094722"} +{"id": "train_08351", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tooling", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "681e6f930847e63b8d9490c3a08cbe504e9865a1"} +{"id": "train_08352", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20853dfd309e63349d481751e9a6d8cee7a1435e"} +{"id": "train_08353", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca7a4a3c04b26b713e6569c3335da9ad3cf8404e"} +{"id": "train_08354", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "390bf1e2c20c39fbfd283dc59a916347d0edad30"} +{"id": "train_08355", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bff37fa029ae668ff8f90cb2394aad01354c97e"} +{"id": "train_08356", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "999877a4824868e2c2d772eac74410ae8d282798"} +{"id": "train_08357", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b607b0ae07c168db74a2279896e5e6c3ac355ac"} +{"id": "train_08358", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c88eb69f0f77bdcd1a2e4c5482dc18bb3fc492b8"} +{"id": "train_08359", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "cost_latency_tradeoffs", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0aa85222695fd7c94896f8461be7e4576070456"} +{"id": "train_08360", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da49f2ef73a859a84b182d697d2a32d185786253"} +{"id": "train_08361", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8291588bda31045f8ff9ae83761ca48621e6ff0b"} +{"id": "train_08362", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50c778d0c00257c3ee118f1416133907fc41a169"} +{"id": "train_08363", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f74d85a5324375b5b5dcf3048cb5b3e28d1e3f4c"} +{"id": "train_08364", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb88abaaa4573e3e3bac719ac61ef424f6249f6d"} +{"id": "train_08365", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "evaluation_metrics", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c63c9102fa395af58cac038b924c0e5fa93cc6de"} +{"id": "train_08366", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54df691a62cbe40d4154263fd0aaba9e7a5e78ca"} +{"id": "train_08367", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "governance", "tooling", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22b1d1b6baef45cf4e3addb98f486bcb887b2e0a"} +{"id": "train_08368", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bd73341c693fd9fb6b0945c86544dfc02324a1e"} +{"id": "train_08369", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cefdc68e2306cfae5718a1a04c1cd278484b296d"} +{"id": "train_08370", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13ae55795de15a792dbca0f9bc849c22ae350c00"} +{"id": "train_08371", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "tooling", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78d990af4c2307b91eddafeccdbf33aa091f6a9b"} +{"id": "train_08372", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "caa222ce71aadd887afe456feff5a2f3264c49a7"} +{"id": "train_08373", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "auditability", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe1664c5f54a1317b92db658ce7b8a05c5353b4e"} +{"id": "train_08374", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "repo_scale_reasoning", "tooling", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3274dcfa40ba487eb383eb650d30207ccb8186b"} +{"id": "train_08375", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d7e13a2025432aba7d95cdabfffec4081f38987"} +{"id": "train_08376", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bb3586a2d8efab3ab6dc1c452696d440240a980"} +{"id": "train_08377", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "documentation", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "206bf5f7240a7dcfbad1e0964c74699551c8d88b"} +{"id": "train_08378", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9856c8d77ceed56f95a3336c558d0f23c1b6be9"} +{"id": "train_08379", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc68e4531542771d4890c22490f2651b459cb968"} +{"id": "train_08380", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9466ddf7c0b3e5e7fb0ecf536cbe5e6676f4f0c9"} +{"id": "train_08381", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fedd81a6a619330a685d9dd52e1a9be9b74bffa0"} +{"id": "train_08382", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "governance", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "095bbf83e46660b80fa7806c41ed2e5a8ead70a9"} +{"id": "train_08383", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ac2b4a05447038b3e4b2095112735ba8d14c986"} +{"id": "train_08384", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "251b8f1e74cbee29037dfa0825e17caeaa6e2f20"} +{"id": "train_08385", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbe2d3c0b4fd59ad0c18190af9e2b81ae74751fe"} +{"id": "train_08386", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dab79b608504e00cb74661b990ae0568d6f641b"} +{"id": "train_08387", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b148079bf0c8d2146b981072d7ca906e173cf81"} +{"id": "train_08388", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c55022c71c2417f4e2943910b4c1f3425639819"} +{"id": "train_08389", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a44d2c6fb3f9cbf6ddd7254e156bd0ad4ba760f0"} +{"id": "train_08390", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "458dd9a67482154c63dc272ed7d90cd1832b1e2d"} +{"id": "train_08391", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "documentation", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e14b28c6d5f73081f090eb8e600eb06fe79a9d0"} +{"id": "train_08392", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ada531731f19b77cc5d157539c3ac4375b86e9dd"} +{"id": "train_08393", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1bb34ade82c2ac09fdb1191eecd0ba51b401ac3"} +{"id": "train_08394", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8046054e4a80752248eb80eedd76fbad8ca5b3a"} +{"id": "train_08395", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "security_gates", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "298d1d795816339b1d4d5f305344621777fed080"} +{"id": "train_08396", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d815a0d2889aba0b84d17bf6c67eb3861a602b71"} +{"id": "train_08397", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e4da2e42b9e9acc994689608107f02e0c77079c"} +{"id": "train_08398", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5aeacd2bbe23905fc594352b04849c173faac668"} +{"id": "train_08399", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eda12ac838e7a900555c369075f596f4d2bef784"} +{"id": "train_08400", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "326dbdbd6ca649c60d571ac1fae099a707172169"} +{"id": "train_08401", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "auditability", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f2560a2f23b0e7cd7c6c423ad44048366d070cb"} +{"id": "train_08402", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b06430161e2317c9009b2f5cf0189e803dd5caac"} +{"id": "train_08403", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c421e633a98a32396bda9711bb78dee4aa493c32"} +{"id": "train_08404", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "ci_integration", "security_gates", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a033be5c25aca095d851163b69f26df67180ebf"} +{"id": "train_08405", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec74a038d41f1a938e1453ee1f7157a395c052ed"} +{"id": "train_08406", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5693c42b7149d5e334d7d30fb0505c5350d73431"} +{"id": "train_08407", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "461df607e26b8404a531eb9a67314cd1ab6b2b76"} +{"id": "train_08408", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "ci_integration", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28de5dab2c841a63dab75ed17e723d29fb11dbbb"} +{"id": "train_08409", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "ci_integration", "tooling", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "548957317b48e1e6b8beaecdc438f4e7f154f63a"} +{"id": "train_08410", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "security_gates", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "762d7511047cbfffae46949873985dcecfbe13a3"} +{"id": "train_08411", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c22e92e7e8634a90dd79b4ae7b2565b98829822a"} +{"id": "train_08412", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff5ab38f25ef15a36a0765aec275edda38dff370"} +{"id": "train_08413", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "documentation", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5c91d672cb6e7f077d012d1a753b8a8b82ada5d"} +{"id": "train_08414", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9987d3d90900bf525ba8a193133a3bbc0f866bb"} +{"id": "train_08415", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "auditability", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b30cd29a83d697669e6019dfd4bf79ab5871731"} +{"id": "train_08416", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a82fd8995d8a78d163fb227ded675001828931cf"} +{"id": "train_08417", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1897734cd20d2ab275c9381cff602ce4c39cc843"} +{"id": "train_08418", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1a44373e98e4d5186b8ea47881c152ed5377b7c"} +{"id": "train_08419", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "documentation", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b396d7d5566a2f9b9e9169760e328b2ff976bca"} +{"id": "train_08420", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "governance", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27ea26ba46e5c1fc96fae1489219082ab3062951"} +{"id": "train_08421", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73f60dcf6ea236f13915295e4e0f4cd73759ba66"} +{"id": "train_08422", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc7771e8ef8e9d8f6309e06df1b82e422593818f"} +{"id": "train_08423", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "auditability", "documentation", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72cde4bb62c0ecfe135382a7ebf4486b79174688"} +{"id": "train_08424", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a25dd7c1ba82b1446d96d854c827dab1e7f4a04"} +{"id": "train_08425", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fb806db2d948a390342f593b91fd7562b673da7"} +{"id": "train_08426", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84e24afa8da546964c8bbdba0043fb446e87161d"} +{"id": "train_08427", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tooling", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b985dbfe4f3173e14a8bac99cc406b891912fb60"} +{"id": "train_08428", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae556a09e86d9170118acc22dd349363428fc783"} +{"id": "train_08429", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "documentation", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e01842931a9b4c42de9acf3746efc6bcb6a9e6c4"} +{"id": "train_08430", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "security_gates", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "234056a10b479c701c24918edbd23453d9a856c0"} +{"id": "train_08431", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58978bacde52b063432a465b482eed17f5f497dc"} +{"id": "train_08432", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c32e3edc89299f42d6e41854d5be2e6f3ceb9a3"} +{"id": "train_08433", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "auditability", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19e46239d9290b8a3536d1881ca5106da12ac7bb"} +{"id": "train_08434", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5bc5f42ccf30e218ff26857b02392ca1c82989e"} +{"id": "train_08435", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "028e45a913e687bf97f0e00d13d329530970a9ee"} +{"id": "train_08436", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "reproducibility", "governance", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c18659a2035ab244e984ba68735996bbaae776be"} +{"id": "train_08437", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfed743a90838fea95180dd8209ddbb13be83fcf"} +{"id": "train_08438", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83ba259994fcb88fa317d74034da3c11977f8f73"} +{"id": "train_08439", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da8591d851adce1d10abcd317348378fb1428c0c"} +{"id": "train_08440", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "tooling", "reproducibility", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98b720333519bc96494b7a9ad524a84d84489909"} +{"id": "train_08441", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbe4d1d275039daa5f56fa043ca359b297627026"} +{"id": "train_08442", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "repo_scale_reasoning", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0e68631065654b259e69c195790834cdca7ce0a"} +{"id": "train_08443", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3717120ab2dc0e36f27149a46592688e4cb1f8ec"} +{"id": "train_08444", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fd2094d92a38c737918753709563a2c97ab0032"} +{"id": "train_08445", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "governance", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4880967ec84cf26cd31fbed45fd38bd58471b818"} +{"id": "train_08446", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a96fd8511d84102206b2d4b21d23dc4653e9bb7"} +{"id": "train_08447", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "documentation", "ci_integration", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31119fc39547d5d1ffe9b50fecd13dad057b9352"} +{"id": "train_08448", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "reproducibility", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "061c30c4f19ac8d011759a2771d8e043b7bfb19b"} +{"id": "train_08449", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06948c0584d565d86e820d59edc272d7f04f8451"} +{"id": "train_08450", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3ae45e19eaf53bfbc2c158e5c1b13cc9bcb2708"} +{"id": "train_08451", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "governance", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e8e31749d19358df370009266c32ebbc55dcbc5"} +{"id": "train_08452", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tooling", "documentation", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5595c1cb220ee5c6c2cfde50a20df11af6ff7726"} +{"id": "train_08453", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a4290c70fd8b825cbd390870e0dec1095c4eb99"} +{"id": "train_08454", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca22fedfd8d65feab9cabac49ad1fab02e7afced"} +{"id": "train_08455", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b300a9edeadc38bd924d134863621226b8e17e9"} +{"id": "train_08456", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dba62d81b4cd81d8977fc81423b091aa68c928d3"} +{"id": "train_08457", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "ci_integration", "governance", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73a2e8dc6478042d7f5c43601999d461462ff75d"} +{"id": "train_08458", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "evaluation_metrics", "security_gates", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab28892111cc6dc7715a3c21ec486598cef13f4b"} +{"id": "train_08459", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eaa0647f35ad3de265709a764f035dca23126211"} +{"id": "train_08460", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "repo_scale_reasoning", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f95cd7d2ebfbbb661e76b6c48dcee89d4f9085ba"} +{"id": "train_08461", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f3085396c5f25ecefaceb4ca4d6d6ce7abed9be"} +{"id": "train_08462", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e31cd7b686876c72a3ed32d7a4f81fe2739ee5f4"} +{"id": "train_08463", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b38bf21315dc5aaf15da3705f5d919c48e29603"} +{"id": "train_08464", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2deddfc88325b4a15fc2888a80192d246104b6df"} +{"id": "train_08465", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "security_gates", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3734e02c7369e3c3d98567e298d6a0776e4a24c7"} +{"id": "train_08466", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d3f491449d7093d4bbb3a8abe4fe59b4b2fc945"} +{"id": "train_08467", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92e31869f7c7568740bb07e0f29c026a96dd4426"} +{"id": "train_08468", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "documentation", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4ab236d59fe5ca4962400fea149049ef195769c"} +{"id": "train_08469", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "governance", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a389d99b18d941231b683d198dcfce8e9d3274a"} +{"id": "train_08470", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee5a26d7815d2efa851f89c3692b4cbd2af94720"} +{"id": "train_08471", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6198e265c405ae699db6efdeb2deb05ca5c3478"} +{"id": "train_08472", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "security_gates", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83d2d06fa75a5324f93a5a9b4b4ceb40bdbc9df5"} +{"id": "train_08473", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22f7f0a900b4e6710e3d6514e1999d3c407d8eb2"} +{"id": "train_08474", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c9f330902de4d870b383beffebf4cdf47e2911c"} +{"id": "train_08475", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9aee70db563c17a8af40a1c4bdb39a0dbd89023"} +{"id": "train_08476", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a3f8d0fcc3b8e015f0cb595cc781b08e6e92b14"} +{"id": "train_08477", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "security_gates", "governance", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24816d0e44a6a4cdb571ab0d5fd13a56c2495903"} +{"id": "train_08478", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40c57190b8a20ee1859a7e70cbc25044366b0ee8"} +{"id": "train_08479", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "governance", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cccb714a472e22489327e47d476fcf5ba3d0e06"} +{"id": "train_08480", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b61078a9336a81f3caac3e6479f873e2b8a79547"} +{"id": "train_08481", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "repo_scale_reasoning", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35ef3a442275611647dd4ad916f85879995fde9b"} +{"id": "train_08482", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdb1ab99a69d5669befe64b2ab91cdf088aad368"} +{"id": "train_08483", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afa9982da51ccad72c0aeee14dfd0e0b921ec562"} +{"id": "train_08484", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c891956b6e00c96c9e0e1f5504cd603cf16f1e08"} +{"id": "train_08485", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f94561b4bb3914f5365ad77d9e9936174ee46ce5"} +{"id": "train_08486", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fad6e1caf392386fcaeddf1531e121b427046937"} +{"id": "train_08487", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37505396e98c4604f347329673a5692b974876bf"} +{"id": "train_08488", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f148c5f1e660df010309f5cf065ed3a0de610877"} +{"id": "train_08489", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee7e598158f00c990c433ac4a72723625191d3a0"} +{"id": "train_08490", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "repo_scale_reasoning", "reproducibility", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c09e0be322fc48add17cd9c0ecfb814c65859352"} +{"id": "train_08491", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "tooling", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "543d51f4a66ddc6d6b78366dc7c36c8bc748c954"} +{"id": "train_08492", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "documentation", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6b33e3e336033e20bb2b61a35ba4a905382e4fb"} +{"id": "train_08493", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b8e452fc9dd21fd161966be49c899f304a1bfa8"} +{"id": "train_08494", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "auditability", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12cfa5859ea5797be010fd811b21d70dda3c962f"} +{"id": "train_08495", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "repo_scale_reasoning", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d201b6ff079f1baf371487c520052dcb086ae96"} +{"id": "train_08496", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "documentation", "security_gates", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92bf72b7564244e2b65ad3b44664d4f1506f08b4"} +{"id": "train_08497", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8948e54f30c06b7b88d77e33af440b59adc57c2"} +{"id": "train_08498", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tooling", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f235d7d390f62de042b1110300be1c108b606864"} +{"id": "train_08499", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3feb1389744c37b074c32fc7d487f0a853081a7"} +{"id": "train_08500", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e6eaeafe30d1364123ff2012fdd14be520733c4"} +{"id": "train_08501", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "tooling", "reproducibility", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1193ce21e5913e70e36db1590dc7ca04f2566f6c"} +{"id": "train_08502", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "acf6fc05c7ad5c96f1f5c0effcb62953ba9a5734"} +{"id": "train_08503", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cefd391273ffa4df92931fdaefef3a69b0733f8"} +{"id": "train_08504", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d160c81e40bb990a876632aa751860372af063b7"} +{"id": "train_08505", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c47c6c3448a00b79daabe455dd9a2d19a3406dc7"} +{"id": "train_08506", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "auditability", "documentation", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b111f5707d3553afe68d04583d122f01f738da1"} +{"id": "train_08507", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1a389580ce94179a394494f620aeba702103549"} +{"id": "train_08508", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "documentation", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c614863007daaf5c65480572a7c47e0a733a163b"} +{"id": "train_08509", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20f7d76152a2ea5b5b7af37d3ff9c9de619653fa"} +{"id": "train_08510", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6998a19a2ea7c54fb040966bd9c7b8b55ae0f0e5"} +{"id": "train_08511", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d964a2d5b9ae9e722ae92c063f189f7bd0d433fb"} +{"id": "train_08512", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "governance", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b635f408caa4c8da3be8f02cde75e3faa5edc6db"} +{"id": "train_08513", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "governance", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e25c4e2e0bf4d9cfecf1896518bf35863edd6dc4"} +{"id": "train_08514", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b259e04d0b1638648477cea1eff2abfc02cff300"} +{"id": "train_08515", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20c3eb62e0a69b71a059c23953ccec564c0b76a6"} +{"id": "train_08516", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c956918ccce99c8a2fe6fb2db0842d93a1e3a432"} +{"id": "train_08517", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "049c22ba1c3e1bb22964a9cf55a21811efe4b911"} +{"id": "train_08518", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12d10c0ca91dd1e073e18d87446eaf553c9b1387"} +{"id": "train_08519", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98eae7fcbe1344ed036bfe31552da3a13e9753e9"} +{"id": "train_08520", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0bd73c5637d677c55f591f8ce76ef041f82440d6"} +{"id": "train_08521", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "evaluation_metrics", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "016f2e27d541a4b56ed5171fe9fbbc211bd42e98"} +{"id": "train_08522", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a3125e09162fc63021379bf73a33782ab960df4"} +{"id": "train_08523", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "documentation", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12429a7ea054f311c6dd3712225dce7853d3193e"} +{"id": "train_08524", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "ci_integration", "security_gates", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e7ac289ee1cfb5830ae554724a17235da2ca5a2"} +{"id": "train_08525", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8c177696d70c81b23e9b0c24f51fc115f023d97"} +{"id": "train_08526", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f1550602e9a44fc9eaf4b3b203b6781aa17a2b2"} +{"id": "train_08527", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee042074a6cfb87dc19cfc6501941ada64539cd6"} +{"id": "train_08528", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "evaluation_metrics", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "299f02e5ec0fbbdc94e82f9ac010748f9433c0f7"} +{"id": "train_08529", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9ce1f3963fcd3d576eac71f51b817b975a2ba50"} +{"id": "train_08530", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bda80f03d010bf1a3575fa3e26f62c4ecc07f264"} +{"id": "train_08531", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69d796a41004677faedb4fea85996b381fe1eee4"} +{"id": "train_08532", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68264a28806318a15ee3ce721b431a1d49444f3d"} +{"id": "train_08533", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tooling", "auditability", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60b1852a943ce8c6b2a42fd63d89313141d45945"} +{"id": "train_08534", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "security_gates", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a832f1e119b4b5bf3b2028a5042e739a4852ed13"} +{"id": "train_08535", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ec3b618a9840ea0e6fa03fcb5e43cc1c4351fc8"} +{"id": "train_08536", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e24280ae09aa4ad2dee3a1c48de8fb95367b341"} +{"id": "train_08537", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b0bad73fc9ad2036fcd41151be1a7017fdd4f1e"} +{"id": "train_08538", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6dc7b18e1aeb6492aea8041d3c74f9cf557670ef"} +{"id": "train_08539", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08ac0be7af5180b962954aa4685c592179a3722f"} +{"id": "train_08540", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28d170fe0324917376e1b92f7cfd776bf5ae0ce7"} +{"id": "train_08541", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "tooling", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "644ef6815088d6e866575c4836d212f097e5c304"} +{"id": "train_08542", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "evaluation_metrics", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f282c9b7207df1ebf780a6a2849069475566b311"} +{"id": "train_08543", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91d3582ada4c08470d9706fa12bea513e737c8e6"} +{"id": "train_08544", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c75fa58178fd0b3c3b18a565095fdb1cbd8e2db"} +{"id": "train_08545", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tooling", "security_gates", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7ebfbbb6cc110f332b1fcc4378bf9e7de0c6201"} +{"id": "train_08546", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40681e7d6060a44a994cff00f0521154aafbc14a"} +{"id": "train_08547", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce93b072bfa41e50ecff4980a1743ab07785c76f"} +{"id": "train_08548", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "ci_integration", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0a8af5567aefa538a4919ace4279696cc42fb40"} +{"id": "train_08549", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tooling", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4a377a20f62c86d569dc326d22e4871c8dfd3c5"} +{"id": "train_08550", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a16461c32833af1b769d9ed025f6484b36d0bfd0"} +{"id": "train_08551", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "reproducibility", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d41742ac6bc3fee5ab93cd19f10ca0d2cf94c3ae"} +{"id": "train_08552", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "ci_integration", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c047e725551bb29efc77cd10c918b7d1b48f6ca"} +{"id": "train_08553", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "061a52b3e523994d45eb335382a5f3ca377a2b7a"} +{"id": "train_08554", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "documentation", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfc9ecd464b96c0f196bddd0c76c4981ec504a5c"} +{"id": "train_08555", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aabc477eb2ca5c772d75b6593ac7b2b74f260b44"} +{"id": "train_08556", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9397d095e22a584736b61ba239696a05ec384d54"} +{"id": "train_08557", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tooling", "governance", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dea26e362a526a480826b9706e72ea46a14d650f"} +{"id": "train_08558", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e893ba5e349de855d2a1f8ed75edb1162b9f449"} +{"id": "train_08559", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0cb83836b70a249b4581176e3868d405ef755b5"} +{"id": "train_08560", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a97dc6d43a98e8a3c5363b89c4d9220e98d4532"} +{"id": "train_08561", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "evaluation_metrics", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca093fc201fe2e1d59f03b5175ac535e24dcfcd8"} +{"id": "train_08562", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a8d43c705c8db205923c12b7e83422eb335309a"} +{"id": "train_08563", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "evaluation_metrics", "ci_integration", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1048071e711606f5afd5361d710ed62f1a30e2f"} +{"id": "train_08564", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76a220e78b5420ef25423366f567a343d523707f"} +{"id": "train_08565", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a200c501fb823f68d1300f2ef1219f5fcfd4b9e"} +{"id": "train_08566", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "936f52cd67e3f3f29bd98570c28ceb5427754fbb"} +{"id": "train_08567", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "governance", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f1027d864ba6adc40b4b63aae28457cdef63384"} +{"id": "train_08568", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6108e9387fb129dd6f501d63faece6064d2074d7"} +{"id": "train_08569", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3962af02e39adea8ec43b5247f59382d67cbf95d"} +{"id": "train_08570", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84d089bbb457346de9b4d8ec4eba1183339ca3ad"} +{"id": "train_08571", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "ci_integration", "documentation", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8572ea98283d08d3fd8bd010bf1f715cee639e58"} +{"id": "train_08572", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "reproducibility", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "686f6c335aeb21a8bd58d2380b2249016e8b4934"} +{"id": "train_08573", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9623e431d6a992bed4928ca8a993154ef4778c8"} +{"id": "train_08574", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "212fd619d9dc946a5baf698982795a332c3523de"} +{"id": "train_08575", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42f212a06aaaab64c2ba0ba61c715798d0910629"} +{"id": "train_08576", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "documentation", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5c04876c3a0da32658605e0f0244a311d736d06"} +{"id": "train_08577", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a12bfde9d1db3a5504960c6826e94fd297024b9"} +{"id": "train_08578", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "auditability", "governance", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbce43d0ba91ed01e1e0d11312daa691ddd5320a"} +{"id": "train_08579", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "governance", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3c6a52e620a321abf62ec9f5e86e9bf0f3efa40"} +{"id": "train_08580", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e23989027bb01cdcb89620b13ca2ad61b850cd31"} +{"id": "train_08581", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "162aeb95ed0c2b4b8f4639ab14644742e1064e4b"} +{"id": "train_08582", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tests_are_truth", "governance", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78bc6bafc3bb1ca9313365c176271fba830d3bd3"} +{"id": "train_08583", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50cb0f89adff092882933244d434124344b24f2e"} +{"id": "train_08584", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58ffc415ac596e5bbaa3f7a3f418d1225cc8ae16"} +{"id": "train_08585", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "reproducibility", "tooling", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7467f2b9083df814e317193d54aab2ea3f28949e"} +{"id": "train_08586", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e7ca7a8807624ea132980515c9edbbad13753da"} +{"id": "train_08587", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "documentation", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cc18ccbcaeafd189d21b53ba750120c3bd34222"} +{"id": "train_08588", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa6e4f8db9f4ffe078acee91304953f0b598e08d"} +{"id": "train_08589", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "security_gates", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88835deeb7b9cae958759be8f14d4083a7235d0d"} +{"id": "train_08590", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b316869ba3d565b157cfaaf088d05d044603ded2"} +{"id": "train_08591", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "190fca347e55c4925a2ea0cf9ec5e491cd386ad8"} +{"id": "train_08592", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56d11a4f5cf81951b7d663cef0fa640ff8e6106b"} +{"id": "train_08593", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb104cd57514bcdc485c743d20dd5d706b6fae5f"} +{"id": "train_08594", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed2d50b371cadb72948c1f4045acd669cac668a7"} +{"id": "train_08595", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50ab90d4d36ccd945e2dba6351df5ca51fa20c7b"} +{"id": "train_08596", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7250651ac0676079cb424d8e9cd19535bd1b2fc2"} +{"id": "train_08597", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27beaaeeab855d2314c36ff885a62c77f792b645"} +{"id": "train_08598", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d6e189b622425bde24b147a07ececdb384b10dc"} +{"id": "train_08599", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1183aa5753977185728191a13afd538d1b8b40f5"} +{"id": "train_08600", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "reproducibility", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b1d0812d8ed89dcbe0a46c9b08d04fb3a29aae0"} +{"id": "train_08601", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "documentation", "reproducibility", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7e252071df823f941fdd65de0a6205fc89757ce"} +{"id": "train_08602", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fae98ff39059acf44bc156306da2d70ac59310c"} +{"id": "train_08603", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c91909099a4157549b0cca6d5a4aaa9367dda2ac"} +{"id": "train_08604", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "ci_integration", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d600d768995d67a7e0b74326d8841776d968a56"} +{"id": "train_08605", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff167e2dd89774f30df63247c7a06cd505951765"} +{"id": "train_08606", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "795698d3569748dea7a71f536cfd28f73d5639e9"} +{"id": "train_08607", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25aa2a7436a8f7c74c89fa201cd1bcf3fc7c2006"} +{"id": "train_08608", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "documentation", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5245c44820611622330bfc2f59f24381ee16791b"} +{"id": "train_08609", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23be13d0276ad9622d48abb3ea3657c9ee9b8414"} +{"id": "train_08610", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "reproducibility", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c4f5469c81337a271999f5c56c0d3e4bfc5439c"} +{"id": "train_08611", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc4e393c192aad7e3ae353dc48063162a75dc22e"} +{"id": "train_08612", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb3d9b3c6ccb2244447f5738a9f670464758e0a1"} +{"id": "train_08613", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8838b9665a90f082a348223bcb12c688bb946153"} +{"id": "train_08614", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d32a3e5e7e1edba8f5815c8d53855c5acc3475b"} +{"id": "train_08615", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0840b66efd5aedcc368991306ee7107c399ce63"} +{"id": "train_08616", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "governance", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bada5367d321c4e6de66a66dac06afddd7441ac5"} +{"id": "train_08617", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "governance", "documentation", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7037aa64f69b49bf12c9e1d3ff53de350606cde2"} +{"id": "train_08618", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "security_gates", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86984e5570e53bfeaab5448927e980f9fc1ee783"} +{"id": "train_08619", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6544cad276c07e214ed4d1eceda9b4597e294b8"} +{"id": "train_08620", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37cce62d41ebefd515d7f286615e32580c925499"} +{"id": "train_08621", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d52f97f544c0e78cb41fd80b011e07fcbbefd951"} +{"id": "train_08622", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e23bbb3c082c9edf0dc183934e8ee6af7ac7ae70"} +{"id": "train_08623", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c535316cc2d413424192ae85ac3b1f0bb51ca93"} +{"id": "train_08624", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "340c31399833acf17249d7c61dbf750dd7672df5"} +{"id": "train_08625", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cf659b223a8c271ac77e90cebdb99d43a43c32d"} +{"id": "train_08626", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4913f029509d6d88cc78b1f7dc260ab7444aed21"} +{"id": "train_08627", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "reproducibility", "auditability", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b50fb4eadce1f8efbf0affbb83d0fc7e4a0a12b"} +{"id": "train_08628", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43ae297fdf9cdc060ae9798439ae1c12dcccbeb5"} +{"id": "train_08629", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tooling", "reproducibility", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1890c89ad7c81e5270f2733a9e32feb99dbab43"} +{"id": "train_08630", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "security_gates", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5547391c52f3c79f1142268fa0b606dcf01210ad"} +{"id": "train_08631", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7666c4d2f9fc36bbbc402a2c89670c4eea8dfe2"} +{"id": "train_08632", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8be6fd34b4d0132f13ce921ec751c73d62878c6"} +{"id": "train_08633", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "evaluation_metrics", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e2dc297d2e68ef49ba057a28b59b5c6a8b85727"} +{"id": "train_08634", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4e4c852428f688db3c8100c09cf100bfe4aea47"} +{"id": "train_08635", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "evaluation_metrics", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b98bb0240f54edc63d3345d5e0f5269d90e60422"} +{"id": "train_08636", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b1d0812d8ed89dcbe0a46c9b08d04fb3a29aae0"} +{"id": "train_08637", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "ci_integration", "governance", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e0f81b4f575c16e1fa34e80afe84021a3718c03"} +{"id": "train_08638", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "059bddf2d9fd74cfc2d87ea846528f24fd9da97e"} +{"id": "train_08639", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "auditability", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c15d62484936b9da7a65aef5af872e7108926d57"} +{"id": "train_08640", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23c1d229f1eec53745d92746bc6ac6752ddd4681"} +{"id": "train_08641", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "603fe480aa500b36094a63da7b7f0097de0a4f31"} +{"id": "train_08642", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cbdbb46984c0489627834d9b17d0adfefb30900"} +{"id": "train_08643", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "auditability", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0597e3c7c40b7b6795f579f6e5f4ea7f3d7bf672"} +{"id": "train_08644", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "security_gates", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96e71c02541f187371acada127ef3d91176addc2"} +{"id": "train_08645", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d684b4827bac8f10a4776445d597bb6b93847f2"} +{"id": "train_08646", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96248cf5e9d92f7d41d6121d743a5b1f19e0cb3c"} +{"id": "train_08647", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cae70778d85910c4788932865934945bfee52650"} +{"id": "train_08648", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7363c627c1f7687b1b5d0b3fb43ea2195803be63"} +{"id": "train_08649", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c2c939d655f1fec4fae390fef08a56aff7109ac"} +{"id": "train_08650", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a8711c515613be5095d04e2a5687e98b0b3dfa8"} +{"id": "train_08651", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40099fa0a216a4718a7f8ccce7883410f44215aa"} +{"id": "train_08652", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b1f9b49ca06ab701fd06e4f356802297f44c74f"} +{"id": "train_08653", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4fc366b93440570e2b2df1d601a2d7d330413a9"} +{"id": "train_08654", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eef057c5d9a2af9b8a7a2261d20c1974ae010ed5"} +{"id": "train_08655", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0cf8dbba184c70066ddb781bbb700aee5d18610"} +{"id": "train_08656", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "tooling", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "058fd4989b5250009e70e95e8ca83d4c527bbb3e"} +{"id": "train_08657", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ef072f04a1743efc00568f6cbba548aa7fec357"} +{"id": "train_08658", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "evaluation_metrics", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6aec82bf6f2abd0ae99fb41d6db090f2a1a28f8b"} +{"id": "train_08659", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "reproducibility", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf4e71d4f16493e21266dd9f46d67d46cb02e6ba"} +{"id": "train_08660", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "275986e0fe7de35b4e0853dcecb374eac0f5d153"} +{"id": "train_08661", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "documentation", "governance", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd2662c2639567378722c510c18f4e38c6fcd7e2"} +{"id": "train_08662", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6dd4bf0446015100513639bdace668d8a38c01e1"} +{"id": "train_08663", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e888c309c790b35565f69d453e0b4328e3943e7c"} +{"id": "train_08664", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a84ddd38c4080e98a656b441770e6f3c7f2898a"} +{"id": "train_08665", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "590109a03440fe0858dc4580b3d874739b9461d2"} +{"id": "train_08666", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96fe001e36bc7613f8909ada150d22c0aa6a3306"} +{"id": "train_08667", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4dfc16351e663393646c7b8904b029397719af2"} +{"id": "train_08668", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "governance", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "429020ec1bc9a558ea75c6f4fa49ae5dee2fff3a"} +{"id": "train_08669", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "ci_integration", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65297764e88bf46e30e735598fd4fdb2cd0c8044"} +{"id": "train_08670", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tooling", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be056c1f9dc4b44cc35f58037b0f3eb4f1ea0e86"} +{"id": "train_08671", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6951825cead18ecdacf07fd47235eecb642dd14b"} +{"id": "train_08672", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57837035adf017306e48aae22c8cf08cee93e368"} +{"id": "train_08673", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7aa0c33a53b0fa105e5813f851aa0007a7d198e"} +{"id": "train_08674", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83d2d06fa75a5324f93a5a9b4b4ceb40bdbc9df5"} +{"id": "train_08675", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "tooling", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f9560f9953f1e15ce4ae13acc0ce785a4066a8a"} +{"id": "train_08676", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba53ef0be181bcd0437802f1b57107c4c04f6f6b"} +{"id": "train_08677", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "documentation", "governance", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "029879b4aad9e0bbc7df5f85971bdc48b7a71832"} +{"id": "train_08678", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "reproducibility", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "638f434b76ba514e4c5d5795b9a843dffb6c9879"} +{"id": "train_08679", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "reproducibility", "governance", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbf4a4f463c5a7b3539328f37105ca6a4e6ec51c"} +{"id": "train_08680", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "repo_scale_reasoning", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddcd471e6318b2fcdd152fc7c0b5d8ee84e3ba01"} +{"id": "train_08681", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "767ac1923594a1f790bdc7cec8a45d5c39dab5aa"} +{"id": "train_08682", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9164a332534fb31d064b309e90248cefc9a50a4"} +{"id": "train_08683", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60f869975050f43ccaf1f59feb3b53d461b0ffca"} +{"id": "train_08684", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "security_gates", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79ce9f41414f5287d48a4d84d9bf8dbbf1c51035"} +{"id": "train_08685", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "reproducibility", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f8dae52150dcfd0b9bf2aad7c8e4ca671b35cc6"} +{"id": "train_08686", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4f826c99e851c51112aa8ae83d81b0c5cd695c0"} +{"id": "train_08687", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "governance", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1aaf526d3bc9be5890181da419a62e54ad784d28"} +{"id": "train_08688", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "security_gates", "auditability", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ded02dd70830ef1194988eb1bb375e8b1c04b4b8"} +{"id": "train_08689", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b3a1b31bb0516e7de9d5ed3a589164011b66dd0"} +{"id": "train_08690", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25aa2250d710a2007491305b9539e6ae31e4a4d6"} +{"id": "train_08691", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6f314f2d60e0ce48bf779e486b039ffe3fbd2cb"} +{"id": "train_08692", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "ci_integration", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fa63c37d1b79399c4b91203efeb6872b08d17b1"} +{"id": "train_08693", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06320d0c0873f0aa05615ef60af63ff8e6f755c4"} +{"id": "train_08694", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b15134e997bba863bcde1d82a75b22777469b114"} +{"id": "train_08695", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d7eb4eea714c2a866f28099e658b6ac04c854da"} +{"id": "train_08696", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92652877065e7abc091e1cb526352dc4c93cf023"} +{"id": "train_08697", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "680da2d1f0de1ba2b2760ca766b22797436a13b1"} +{"id": "train_08698", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9b580f79510c4f8d2a2c89322f985e435f624f2"} +{"id": "train_08699", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "914565862c40b60c4046fa04b2c4981fdab6b90c"} +{"id": "train_08700", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9efa1a9a772332e130f6d88ea946cfb361c36f36"} +{"id": "train_08701", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae95c77ece7c20695ef27c2b6b84695ae923313a"} +{"id": "train_08702", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e154c1dc7609f1eda62ea43220a8a56650c7008a"} +{"id": "train_08703", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cc392c4f3b1fdec7b2fa01ea2a98b0e2ec3c6a8"} +{"id": "train_08704", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab3d20cf307e104a7684b55fab3e9c0d6f8c8475"} +{"id": "train_08705", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "531ac2875c18ab596a191e67eda0add253a313cd"} +{"id": "train_08706", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8a47e82babe59f13035a119e38591fda61171dd"} +{"id": "train_08707", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "repo_scale_reasoning", "reproducibility", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59aab707a30e2584476a6ef30823ca999ae4bb2b"} +{"id": "train_08708", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d62bb4975e7e55e0fe554acfd2d5944a89c44099"} +{"id": "train_08709", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c68a9e5bfc00f2cfd4b383489d7ac3c12feaa4a2"} +{"id": "train_08710", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "544bc27f1f5222ffb8509ecb7a2a6b6f4c906afa"} +{"id": "train_08711", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f157b4345055db126aff29380cdb1b7eed1c0cd"} +{"id": "train_08712", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3d83dbbb1321a3fa569addddee5371c4c821614"} +{"id": "train_08713", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "security_gates", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b025663d8e832fa6e7363f6d1ddd4f7f424e25d"} +{"id": "train_08714", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "cost_latency_tradeoffs", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "abb3da8d765df5f6b947c70baef2d7bdae47f205"} +{"id": "train_08715", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1356476f063ba9efb9971401c566e348b2a7ccd7"} +{"id": "train_08716", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "086fb378f84a67d5449906135e62835ae3089dbd"} +{"id": "train_08717", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9aca4dbbd26e307b1d0aff8ec3294505bb74354"} +{"id": "train_08718", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tests_are_truth", "governance", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19593c1195bf41473cdcf60bca41e20869a6784b"} +{"id": "train_08719", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5161737a6bfa3296971efb8e0f085fbff2dfe53c"} +{"id": "train_08720", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "839ed65cdbc1c25dbcedf60101d9b1ed87f68857"} +{"id": "train_08721", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a8ec48ab57de057932229107e14c4f4b864772c"} +{"id": "train_08722", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7321675f12369406eca0b95ed4c6caac605ee6ba"} +{"id": "train_08723", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65a03680d08dbfbe8e8ab25c368180ecb7c4a962"} +{"id": "train_08724", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5c845ee2a1af5669fc7f19c01ded74786a311b9"} +{"id": "train_08725", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6058e65753cda90057ece1e6c1322e0bedf94eb"} +{"id": "train_08726", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2d355d876419d0742b2f1413956d51194db6e45"} +{"id": "train_08727", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tooling", "security_gates", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ef523b4b30c563165e0b9cec9f1564f71c9ef74"} +{"id": "train_08728", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "documentation", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c689137c0dcd8b5540af1550fa00a46b5f4adf2"} +{"id": "train_08729", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f123ab618917bd1acd247493b07257afd64ee81"} +{"id": "train_08730", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ee1e0b444da60df9208d10b361605b76cc573b9"} +{"id": "train_08731", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0eb8a06942b0211868595ddc62fff87bc210d941"} +{"id": "train_08732", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d75807c6754b6609de38d0a0b011eb0bd914e1e7"} +{"id": "train_08733", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "118c5013e7886f32c7f2f82977ca39d14239da49"} +{"id": "train_08734", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e230847e1356ed26fc97c6607de4fcf648703e9"} +{"id": "train_08735", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2145a6aa934e3d3bc73e78ce546d21e60abdb317"} +{"id": "train_08736", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "reproducibility", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39aa123dcf35617001be9c6df31e78f804bcb67d"} +{"id": "train_08737", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "ci_integration", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "742bc4b3237368787e7c126d55ac9268aeacfa60"} +{"id": "train_08738", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "reproducibility", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6776a724ec6a6758eb9d24ab1c003f511b8a2d66"} +{"id": "train_08739", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tests_are_truth", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f52e68d77a996c568f64dacb3cc71df0e58f3fb"} +{"id": "train_08740", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "tooling", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7091b5ab7717db22b189f9e6d0a7d01b831ae15a"} +{"id": "train_08741", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "tooling", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f994de33a8463326ad02ae4e9abd7ede84dc7fbc"} +{"id": "train_08742", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53d847bf7cb7b09b1d3a6596988f54eb84266e49"} +{"id": "train_08743", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bfbd1941e47de9a3cf87be24a7f3154c4bf02df"} +{"id": "train_08744", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "reproducibility", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4866435fe9d945a093c55fabd1f144bad7715c66"} +{"id": "train_08745", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "documentation", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e3f8403554c3facd6bb536fbb4b9e210a200963"} +{"id": "train_08746", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df02baebe799ba92785a791e91f702dcc906f153"} +{"id": "train_08747", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "documentation", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b464b89e9944f0f157397f195eda20614f07e62"} +{"id": "train_08748", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b540168179dc88a19e5ea72b5c9a789e325dff96"} +{"id": "train_08749", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "ci_integration", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b23c16c2cf9e8fe970ebc57765c27cfa873f9315"} +{"id": "train_08750", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "governance", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07de88a0283d234fafc08c97d840996f2bb3e5ca"} +{"id": "train_08751", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "auditability", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc4f8b9e0d4f8a4e6ef9953098cb92a546b1a7a1"} +{"id": "train_08752", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8773555d872a340112d63103397d36f3eeca886e"} +{"id": "train_08753", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "reproducibility", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d1888160012f08d137d146e62ea1fc89ff12abd"} +{"id": "train_08754", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3cdfcfc0a3cc8690acf1275e266182d4fd61d96"} +{"id": "train_08755", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d006ac430b5e1d79f0f92222ea12ea077846f78e"} +{"id": "train_08756", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "tooling", "auditability", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "109c785f79f46191cfa100e78015c2e94671f0eb"} +{"id": "train_08757", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "ci_integration", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eec6f43c8bc3cfbeb00df1198068ebcc837ca1d1"} +{"id": "train_08758", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "documentation", "governance", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3738671ead2fbfbdc65ab3ca601bb39977d512d6"} +{"id": "train_08759", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4a3c88187e7ddd2dc9c6c7d2dd2bf901b1ed39b"} +{"id": "train_08760", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3758388bf2aa41c107006c3837622bea24ba45a"} +{"id": "train_08761", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7edf9e9e29a18dd599f4d672d84150ac62a291a6"} +{"id": "train_08762", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37394e6c6d65a69b04d3d3a93455037f60640dc5"} +{"id": "train_08763", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "reproducibility", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f926d59708b3944153b23db9b9e1fba5fdd80552"} +{"id": "train_08764", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tooling", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cad2b27858649acda92b212a18920980c8b29c57"} +{"id": "train_08765", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18c1892a74c9b36a973420ac574c83684105b5f5"} +{"id": "train_08766", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7ff266eb60b027d67395bbab8b0e7632e6757ac"} +{"id": "train_08767", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "security_gates", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d605896deecedb73debde6038a998e80e023993"} +{"id": "train_08768", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "983a7d1339c8a3cfe887539c97d98d8b146fa235"} +{"id": "train_08769", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cd4e8b3e35fce4fbe4980d6c698f1a887bafad9"} +{"id": "train_08770", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "governance", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82767d20dc5ebae1b1745ba1bf74c1d8c9b88e75"} +{"id": "train_08771", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e3d21bb280821b8f865b506bb071f392fe1a227"} +{"id": "train_08772", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "379eb6b2ba9932c9c10a6da5097bdbb239dbb9d7"} +{"id": "train_08773", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1d09aee2c4c43a08632cc47291363a67a06ab18"} +{"id": "train_08774", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81a0a13370f4d067990eeb94ae3261899f1e16f8"} +{"id": "train_08775", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dbe2c32ab0f07e3afdd44ffada0b27d48a866f0a"} +{"id": "train_08776", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16ed11c2e575d4920e9ad4011dc04b2afe44817d"} +{"id": "train_08777", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0d304406c175f82e972b97846df9265dde4bbb1"} +{"id": "train_08778", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "security_gates", "auditability", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58e36ef314bc7d3d57365693253ec3cb49349a1e"} +{"id": "train_08779", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ce80f631e6fca3491162d4763c32001e3761d8a"} +{"id": "train_08780", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6b618306205e6534d285fbd95b71942c39acf80"} +{"id": "train_08781", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e0eadbbb6b91bb69fe663ca569190f1b6e8cd1d"} +{"id": "train_08782", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73b7a5c735f06fb9a3e78ebf99463c6c9a99d22c"} +{"id": "train_08783", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "528bb2d1a7ba2c25709825b19fd6d3a60ede2f2f"} +{"id": "train_08784", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f394b9bfb36f5ec399857b242ab10bf5ae689c62"} +{"id": "train_08785", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "reproducibility", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f23ecb5e6357c5f1ea81f216f617dea2c2084701"} +{"id": "train_08786", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5291c4b98de3240c350177685feb2f044103797"} +{"id": "train_08787", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c95b593fc1d6dcafcdb6740e8c9f95669cda67f"} +{"id": "train_08788", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3eb0a585bdf23ada423d04e42c7155bdbb725592"} +{"id": "train_08789", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "security_gates", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2016148a7bb1a6574eddf1f0b1cfcc26bb3e3442"} +{"id": "train_08790", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "governance", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "156fac7dae2f81d8278175baf392452462bfd9ab"} +{"id": "train_08791", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dbca21f4a692bcfe7ad8deae23b4166ed100ed68"} +{"id": "train_08792", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9acac950b622f8ade24a8b56fe52ef55c3a153c6"} +{"id": "train_08793", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31119fc39547d5d1ffe9b50fecd13dad057b9352"} +{"id": "train_08794", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6f2ddb21203fb57a814a907dee642bfdb02891b"} +{"id": "train_08795", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bea23bd133244b16b8899bb89e01b7b02cc84366"} +{"id": "train_08796", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d963016ee80bab3c7cd82eb1b5cab04c06b380d"} +{"id": "train_08797", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5781997bf00a71a09a40e51836495025f185a9d6"} +{"id": "train_08798", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e629f46f4ec2b4026b197b15333e68c6fbef4b1"} +{"id": "train_08799", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17fba26c3dbf5831ad085aeb1a71ea0555418fdf"} +{"id": "train_08800", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af80e041f7f3e52d03a0963bb5f9cd106ddb3ae7"} +{"id": "train_08801", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8be5c1359144136dd30c366c53b30abfffe0a337"} +{"id": "train_08802", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "314ed10ff5485d1b4a7945291aa669e4608bc3f4"} +{"id": "train_08803", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1415d596d5c74eb88b319035fa571a6312f20b13"} +{"id": "train_08804", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3556719c9655b93b9cacf04ae19fa82b4ae9cda4"} +{"id": "train_08805", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "tooling", "documentation", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "230073e1ba3353e8b2a8eeec3a0a26be6bb8bba3"} +{"id": "train_08806", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d34db206b4a9c4c2b26e1609c635cba33e178a0"} +{"id": "train_08807", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "reproducibility", "auditability", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a490765279bb9f6a4f9d63e5d28b021052cad560"} +{"id": "train_08808", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "ci_integration", "tooling", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2059964c637caf2d68cb6c1051efcfe3cd2a4078"} +{"id": "train_08809", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c92dac616e663a45d891970641ca850743411798"} +{"id": "train_08810", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cdeacb641dfec773c826282143d5cf2266fc75e"} +{"id": "train_08811", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58059582143b6c3d085c6bb7bd8d1e015737f384"} +{"id": "train_08812", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b360361ef83ab5ddefdbefa10b8e83892d0f42a5"} +{"id": "train_08813", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "documentation", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c0e96de117514013a45e883e5afcc4a549277fa"} +{"id": "train_08814", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "security_gates", "auditability", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7105e3cdf06d307d66da4a099449dae367ca0478"} +{"id": "train_08815", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33b0416fcb421ffdb2b4b78f0d9036925b6bf88f"} +{"id": "train_08816", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "auditability", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7206e8255ed0eef9d5fb437b3b4dc055660f97d0"} +{"id": "train_08817", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66bc4bb7d9e7c60ed7661baa7298afb4a27fc24d"} +{"id": "train_08818", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "governance", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e7ac289ee1cfb5830ae554724a17235da2ca5a2"} +{"id": "train_08819", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "tooling", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7de73867c10f7b737026ef2f2ca3964a4b5d18aa"} +{"id": "train_08820", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tooling", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57f1393897cc9c9ff44a3149cbf30ee2d39b853f"} +{"id": "train_08821", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf7f2e0f2ab28fa41c72a12945b85c193cefbdf0"} +{"id": "train_08822", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tests_are_truth", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67a1bfa8f6b1f8c511155f23b22c8d425e981e85"} +{"id": "train_08823", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4cab57c0444e2ea77cbe3de6d534003040dfa46"} +{"id": "train_08824", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0593b036da998ceb3cdf6f38229f3550b6522db"} +{"id": "train_08825", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tooling", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "780456c6eb0b18f50df8e5e0aaaecc2601550dd4"} +{"id": "train_08826", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4af52c3fd93010ec005dd5f95910efa89981cbe1"} +{"id": "train_08827", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "tooling", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e14af08fc4ddf48398753d9ad17c34007f720d8b"} +{"id": "train_08828", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "tooling", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bed5c88869957449ec7948b589f5e6c21a5a8161"} +{"id": "train_08829", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7c920d4b9076f3cdc69283a456cde4b4daf51eb"} +{"id": "train_08830", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ce1d916a23344afe42a119efdebe396c2229b23"} +{"id": "train_08831", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "auditability", "documentation", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b3ff8c849ee3f222963b1eb7af99b4215398922"} +{"id": "train_08832", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c925e18cacb8d6ce95c9647b278ed065c60401fa"} +{"id": "train_08833", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1df77f5a391707045d647b3b3350ae06abb20b1"} +{"id": "train_08834", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7542f51d3ce8145be64b5a8836c12cb4b4d8d2a"} +{"id": "train_08835", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03fb2111c0beefdf3aac5b3dd5e769c4280118cd"} +{"id": "train_08836", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "102b5b7a247c6185580f3764257368202ab17ad4"} +{"id": "train_08837", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "documentation", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9176ef0116c3630adfcfc1c60d3a73c42e52eaa"} +{"id": "train_08838", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "security_gates", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85f992950c131dd1392134f717c80ba904d7e0d2"} +{"id": "train_08839", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5fd72752b20feb7d0f3cf3a1b937963cc3a0637"} +{"id": "train_08840", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5acb3ed40a7131e9897c5b65ffc6937bdaa8e23"} +{"id": "train_08841", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70838dd6d7d8b6a246b8207cfb1cf98ecc352318"} +{"id": "train_08842", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "governance", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3309106fa438ea271db0bd0c2c11ab521a98b40"} +{"id": "train_08843", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8112abec3e76860411f093ddabed2c4f5bf74bf"} +{"id": "train_08844", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "documentation", "tooling", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7058c438585d7d81184d3a2d5fb96a4024b061f"} +{"id": "train_08845", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5255804885dad72b9dbc230b510fbfbee2393f7"} +{"id": "train_08846", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f413e73d7821a0d3e2557a3e4d8b029e21a7224e"} +{"id": "train_08847", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d2a96fb5118001fb31dcb020212e265922cef80"} +{"id": "train_08848", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "evaluation_metrics", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54d061d3a51e8216f9de90628926f8151e26a594"} +{"id": "train_08849", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "governance", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26d89b049c843d1cd0504067a60c2345706ee631"} +{"id": "train_08850", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "004f5c33b969a1d02562396593ad69aa986b6bab"} +{"id": "train_08851", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "reproducibility", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65dc360fa88df3074e19d918a0802b9db40fbc6a"} +{"id": "train_08852", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "auditability", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0708930087b4e77ae2439ad3cc469b4a227b8e09"} +{"id": "train_08853", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec95b9307b7375c49f572d5491246d006303fffd"} +{"id": "train_08854", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdc173b23a1b83f706773cfe3a9dfe82d9684ec5"} +{"id": "train_08855", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "tooling", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55891841faeae5dae74f76510ab278d64317aeea"} +{"id": "train_08856", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db324795d01aeef14cbef4454059195cce40c38e"} +{"id": "train_08857", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "documentation", "auditability", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9365f13e7de68b932a101a20a700934bd4b1b5b9"} +{"id": "train_08858", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "repo_scale_reasoning", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "554a165ebd8fd2e44f095a041e8e77beea5df4ae"} +{"id": "train_08859", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49a8b50428f0d6145c82e122e39d73d75adafd01"} +{"id": "train_08860", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "609433d1b8c1736de7484ef370bb51557c66f754"} +{"id": "train_08861", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "governance", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4694e675c2f1849f9b5365bfdec829a6fb0157b2"} +{"id": "train_08862", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "auditability", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98903f42fbda2bbb060901d1476e0bc52d9414aa"} +{"id": "train_08863", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33da635469587a13743e3caa4412257381ab8b87"} +{"id": "train_08864", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db190b7a948bc71442d4d17d77c95a9f3d1ded66"} +{"id": "train_08865", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "documentation", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b3d396a1f79e41bf3620bf515f19db51dcb7d9f"} +{"id": "train_08866", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9812d9bfedf8a8d7d429c750197d5c47aa5e53c"} +{"id": "train_08867", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39f85ead6ff16bb3772f87d3c3d66c2a3f7b5fe9"} +{"id": "train_08868", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "security_gates", "auditability", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7db73e10c6ae2686d9888f0f2ce7e078d3b5485"} +{"id": "train_08869", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "ci_integration", "tooling", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2fff2356287a1d9dae7f8e8df91bdfe60d37322"} +{"id": "train_08870", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61066ebeeaf48860a9b36159f1a8329be22e0d21"} +{"id": "train_08871", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "repo_scale_reasoning", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b24a4b8639ea91c2d995f624c2704f8a0fb80402"} +{"id": "train_08872", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tests_are_truth", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c4aa5be41d96c8593f8955eac3093402def6a8b"} +{"id": "train_08873", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a40a49cc03bb4aeba7185f6c9d791035aef1b164"} +{"id": "train_08874", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8a23a4050ee782d8a3539acfcf807d94fb7820e"} +{"id": "train_08875", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "374f3e6ed93cdadd4f806c6d5d4d8874596dfa9f"} +{"id": "train_08876", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tooling", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a19b8ce6188900e876a5b4120b88c95558fe27e"} +{"id": "train_08877", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "034af03e5a6cdca891039ef8d9547b173d931cab"} +{"id": "train_08878", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdc3032216d4a5ba022260ecd5075a3621586ec7"} +{"id": "train_08879", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2be96933116794d2617ac1ba8b9d58768fc47d65"} +{"id": "train_08880", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8444e6e6e9d916333b9ceb22fdcbbe8d8cb90e94"} +{"id": "train_08881", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0bd5cc366d0a5f066b9ed95dd90b6fe75b1f4f4c"} +{"id": "train_08882", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3be35998aa1177c74ed0ca3ba3834c2c18a45e4"} +{"id": "train_08883", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e64f74928db65a124cf00640ea9ee30b46f1df3b"} +{"id": "train_08884", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "documentation", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a31f737836b634c2dc8f11e605c18922fea954fc"} +{"id": "train_08885", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "881709ce6edf1ec4bc57821e77801654b0e44853"} +{"id": "train_08886", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "568f52ccb5e4f4a08e3223f3e56c2fad0de22520"} +{"id": "train_08887", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "862510931fd7dc629654679fdcc7ee26bbde4a5f"} +{"id": "train_08888", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3177de63b9e645fbc64b4b80b53666f3ed86e51a"} +{"id": "train_08889", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8abf18769a90a1ee11138e46f6d20f8291ce245"} +{"id": "train_08890", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cf57616840bea41181ed955f5b000ba94770239"} +{"id": "train_08891", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b60acde04ac653389b01402d11cc5ea8f6531c7b"} +{"id": "train_08892", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca20c6d9accf22bbba98cd90d84313e4f486bb28"} +{"id": "train_08893", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "750f1b84c950bc7efe53b6b990e92dc6201b42e6"} +{"id": "train_08894", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "ci_integration", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d216768a4981f25c42cba2dacb5b6c25026245c8"} +{"id": "train_08895", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3ee8ebea7e5d42f8669630dd35804c0b77014c9"} +{"id": "train_08896", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2be8cff0ecb2e6ff20385a8c3302d7f864503503"} +{"id": "train_08897", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "documentation", "auditability", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9fa2bf8c871332b712fa8c5902146adb0014338"} +{"id": "train_08898", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19c8ae1aa49b78b86f265f1d590f7be508a83c4a"} +{"id": "train_08899", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a26939e23665523e248babb765c6002c589795fc"} +{"id": "train_08900", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "reproducibility", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b74647d28e07030c298417115b9bfe7ca6c23f38"} +{"id": "train_08901", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "ci_integration", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a842f29ad8b1dee40b94dedfc1cbe085ceb784e"} +{"id": "train_08902", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16e1d0d5f08957d093d29e6d03fd13659dce77ec"} +{"id": "train_08903", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92766120dcdb11312b144d9d42632d3a601fe26d"} +{"id": "train_08904", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "auditability", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a3f424f324686c210cfacb4e5a9a2e656822de9"} +{"id": "train_08905", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed08b45c23b7690f79c4efffc4c0cf10126e82e3"} +{"id": "train_08906", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "security_gates", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5cd725038e9be7984fdfb22ddf5f14a6e82bff9"} +{"id": "train_08907", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa8b48f5408f68e851662a7e46b0c48a66b46157"} +{"id": "train_08908", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "governance", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d595907e9f47bb399886b063b300be21a841cfb1"} +{"id": "train_08909", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef1263083145c2daca406b13fb9e4609f259aeb"} +{"id": "train_08910", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "documentation", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2e4c4953e4c2ac7282b284ce434beb85cf61c3b"} +{"id": "train_08911", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce0966d9135f7add8dcfd0772aec23326e833d7e"} +{"id": "train_08912", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72aa1c499c543c9e51f533f97d2d93a3677c5946"} +{"id": "train_08913", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "security_gates", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe8ae829990c97fa904c86b5cbbe795594ee40ed"} +{"id": "train_08914", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5f4bc994779225cb34a9a6b3757769a6530a13d"} +{"id": "train_08915", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "reproducibility", "tooling", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93782faeb9bdc8cfb0815d1989fea0f794dcd0df"} +{"id": "train_08916", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "governance", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "719b10d57c319e0abc6c81a4d2d61dee0c50822d"} +{"id": "train_08917", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec413fbe84ca6ba25c1db21bb8871eabda05da4d"} +{"id": "train_08918", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "security_gates", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af102255028f89d052f4d8429e9784e3aa6eb932"} +{"id": "train_08919", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "governance", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc36297859c1bc8826152ce9bb04f0db0bc86537"} +{"id": "train_08920", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bea5cb8420ed24512f1485dc6c821b5934fb3eb1"} +{"id": "train_08921", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tooling", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f62943af080fa7ca69791652a6ae4c4db7344bd0"} +{"id": "train_08922", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "evaluation_metrics", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a28b345e3c712f544ca7d62453c98580642d8f94"} +{"id": "train_08923", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "156f72bcba5a58480429457df6da3a239cd06f1c"} +{"id": "train_08924", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "adee0da8942fbc235ac9325c8f4fd77918e1829d"} +{"id": "train_08925", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "422299fbd10d02da0d7ce7de3c9d6f0d9ce235f8"} +{"id": "train_08926", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "tests_are_truth", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b1f9b49ca06ab701fd06e4f356802297f44c74f"} +{"id": "train_08927", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67af70d06b292b3d211f158d1825a8dafe137bf8"} +{"id": "train_08928", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d42c926ae32f8e9eeee3d0366e9a4a8308acb60f"} +{"id": "train_08929", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3560ad61045626f4cd71fb18a32848693b1bcb77"} +{"id": "train_08930", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9475e25da661ce8ffa699fe30aabe37871011cd3"} +{"id": "train_08931", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "governance", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18862ac3ad958053b194c3836badd4dfe0a2d2c2"} +{"id": "train_08932", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27ee7f625efb4a1c58c85f5d44b7cf43f99a5be4"} +{"id": "train_08933", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdf055ad7820d34f90463d1af3b0b3474ac5929e"} +{"id": "train_08934", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62cb8b9a9bbd185da8f87869018dc95dea179223"} +{"id": "train_08935", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb2520742dd4fbf181de9788214994050bce844b"} +{"id": "train_08936", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "498a9d816c69025ffcf443fe9e044b44a0d31c0e"} +{"id": "train_08937", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "138b692478541bf493ea80ad53ea4fd5751af59e"} +{"id": "train_08938", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "documentation", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d5b19930f3c6fd9d6a6b0f283e8e96567d081a9"} +{"id": "train_08939", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1c31949bd9519f5d8c46f7ed63a1977fbe3e1eb"} +{"id": "train_08940", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "security_gates", "ci_integration", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa949f6799d94773ff16ec92ccbc3f1b47ceeb25"} +{"id": "train_08941", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2d66175c49d98f2491e806b365fd77ff0662b24"} +{"id": "train_08942", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tooling", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d59c64b3a45e3aea70bba100543df54a962db972"} +{"id": "train_08943", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ebb4479628650fafa6d6580d390aafb78d47d05"} +{"id": "train_08944", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfbdfb8f4c934b34dc7d705e240759a2c64c49a7"} +{"id": "train_08945", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70c460cc04e89b6aed43967c097f4bbb8b264178"} +{"id": "train_08946", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "governance", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8045f617505289666a97a52789150bd59da2a25b"} +{"id": "train_08947", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "evaluation_metrics", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aab70eba5378d8ea5cf8e87556a44fef351abd8a"} +{"id": "train_08948", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef16eb0323390cdcc32b49ef72c7ebfc4d364801"} +{"id": "train_08949", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05c64e7683bdf194de1755f987c8ce54d62f345c"} +{"id": "train_08950", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66d9158d7cc8b778fbae336c0f6b5fb08a4a3038"} +{"id": "train_08951", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "repo_scale_reasoning", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1c31949bd9519f5d8c46f7ed63a1977fbe3e1eb"} +{"id": "train_08952", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "security_gates", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "985835b9947e453983b92422d6eca8e43d6b1d29"} +{"id": "train_08953", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "422299fbd10d02da0d7ce7de3c9d6f0d9ce235f8"} +{"id": "train_08954", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9bda24daf4db06bc08a5b983491ec4f51d2c6379"} +{"id": "train_08955", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "tooling", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "886e4ed0594e7729598eb43c309deb90689088ea"} +{"id": "train_08956", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "reproducibility", "documentation", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "243893d38ffcee49347a8e012b134f3e63b388b5"} +{"id": "train_08957", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b3c64d5681e1afc542c66785e577e4d5ad09b8c"} +{"id": "train_08958", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "documentation", "tooling", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "133c748f2a54b9f5f3fac22859b574ebc907b82a"} +{"id": "train_08959", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b64891b0482db718510eec8a8c956978cc297f7"} +{"id": "train_08960", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22b1d1b6baef45cf4e3addb98f486bcb887b2e0a"} +{"id": "train_08961", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "077a24815cf57ac380c4646a7762fb10a6318869"} +{"id": "train_08962", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a2099dfd1a4e6a29d066c33962c7693d9b07fb1"} +{"id": "train_08963", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "documentation", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05c6f0212669c0ab86983a6c0b1f231f5bb076ce"} +{"id": "train_08964", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "governance", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f74729b154621a0b53c94b6cc8cec3166aab3458"} +{"id": "train_08965", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "tooling", "governance", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7dff38ae54a170fca260bba47004a3e49cfe240"} +{"id": "train_08966", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "031059f6f20430d1b82b45fa4f8d81187853078e"} +{"id": "train_08967", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d33de789004b348a66d1e8a4070ded90fe841275"} +{"id": "train_08968", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "repo_scale_reasoning", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc9d581b8f4fd479a93f412f580aa10ce945cdb3"} +{"id": "train_08969", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "repo_scale_reasoning", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f45c15c371664418cd47c3baa6b733255b66a6f"} +{"id": "train_08970", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da0dc4c8b4fca87099e6fe5e44380afe224afaa1"} +{"id": "train_08971", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5428ccdcdf6ca69a0e9e2d869c740f4f623ff8f0"} +{"id": "train_08972", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "governance", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "230073e1ba3353e8b2a8eeec3a0a26be6bb8bba3"} +{"id": "train_08973", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcdc9f134df909aab28d9bdfe5e3dba63d56ac83"} +{"id": "train_08974", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5661dbc5a47e63f243ec8e9f4476925867f657d"} +{"id": "train_08975", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "707b216cb19e7259cff2947d8e9340aa14b3ef87"} +{"id": "train_08976", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "tooling", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5af878ee605c1dd68680110170d96a31ac774a62"} +{"id": "train_08977", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a98577db8acd1f9e340f15bd41eaac52a3d958a2"} +{"id": "train_08978", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tooling", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21f4b878bc127de3e01bd435ad4c656ebef98006"} +{"id": "train_08979", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e16547fd3b4037f9654e375aa4f656a3cf76b554"} +{"id": "train_08980", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d79150af6226605b0ed2560ffd6a89d0453b7b1"} +{"id": "train_08981", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a36d5375b64a9686abd0dc5dd8868e30cc029005"} +{"id": "train_08982", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82a38fb9fb3aacb8907882576eeff808abc89b16"} +{"id": "train_08983", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3dfd6461e28d1036b79d15a2c05b749cf2b6d332"} +{"id": "train_08984", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28ffd172ccab05a12512e2a3bbf89b1c1eec71d2"} +{"id": "train_08985", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "governance", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6137af467f680dbc4ba5f8bfcb8f4c33f5c7881"} +{"id": "train_08986", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3993a734a39be74ba0f600c7601aed0dcc3e432"} +{"id": "train_08987", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "995920042f87542470b1e268608202be8e2e86c3"} +{"id": "train_08988", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "evaluation_metrics", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7a3031d97220ffdfdd40e4c7032768a14ac83d1"} +{"id": "train_08989", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46b8cd9ce65fafb769edca560d1123ceb86b91b7"} +{"id": "train_08990", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "514aabf2bab4880ad12c3bc7c9bfdcb7b0137682"} +{"id": "train_08991", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tooling", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed06e6468e432347eb4d979c2a735533fd01b2b8"} +{"id": "train_08992", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "cost_latency_tradeoffs", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba75176db8481219c6969804fc1b1cea4a6dce93"} +{"id": "train_08993", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "482d9cc967a5428bcf39dad8c71a8ae955682328"} +{"id": "train_08994", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "repo_scale_reasoning", "tooling", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40b58efda403c3d1396cc69d7af272ded35fe1d7"} +{"id": "train_08995", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfcaf467c3724658671fe9ad030f05232940bdc1"} +{"id": "train_08996", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b1d2cfd8ded336aa279b282cd9ad34d669bd896"} +{"id": "train_08997", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5fdc30e5fd03f677ad245eb64f6c2eb34d62b57"} +{"id": "train_08998", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tooling", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ae0b7025856f69407a1c6869e65eb44485d8933"} +{"id": "train_08999", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "auditability", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bc5725ef56568a3f6d0cc391a36cb0a3e8bc145"} +{"id": "train_09000", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "governance", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71a417e10382fb493c8242ba152889f838662a27"} +{"id": "train_09001", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c92dac616e663a45d891970641ca850743411798"} +{"id": "train_09002", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c6447f1b5d1d8e3d6934f8afc945a32793711c0"} +{"id": "train_09003", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "198dd560bda3e99b971716f20c5e4eaef7a406d0"} +{"id": "train_09004", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ac8a8df5aa8659606bcd94700d25400c1a290f2"} +{"id": "train_09005", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "governance", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb48a5a771e86f4e725b495f468fae4cfbfa1e7d"} +{"id": "train_09006", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9592bf89a75728e4d0d8e4e05053035dc8611a1"} +{"id": "train_09007", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "governance", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "950a87d7f9dd5bb8e654dfd720c332fbd9b6840d"} +{"id": "train_09008", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c88eb69f0f77bdcd1a2e4c5482dc18bb3fc492b8"} +{"id": "train_09009", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84d5a7911f92a5ec5372e6a9d89c889b7cddc5d6"} +{"id": "train_09010", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74b54fa647bac83b31b50962d4bf2e6ec80007e4"} +{"id": "train_09011", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "ci_integration", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9762443951d03bb05c2ce971e259b7d96bf59d05"} +{"id": "train_09012", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f3b82c3b46ce5151983fdf71c7c46aa7b245ed4"} +{"id": "train_09013", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35a7087b319f9c061060a2ea81060d661462253c"} +{"id": "train_09014", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "documentation", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89580e238df2034ab23fb67305313727dcdcf58c"} +{"id": "train_09015", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8e71c0ddac498dcb1ab6e47e7764e441377a687"} +{"id": "train_09016", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "291ea89d0657822bdc52d2476fc1978c6059750f"} +{"id": "train_09017", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe911b911aea2b9c37769557aef76b674c52d599"} +{"id": "train_09018", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tooling", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "948daeb0f7d9474a690854bb0224695261033fd6"} +{"id": "train_09019", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90425b9bb2887d3df59e98d345d3fa06bd3599ac"} +{"id": "train_09020", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "governance", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af80e041f7f3e52d03a0963bb5f9cd106ddb3ae7"} +{"id": "train_09021", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "ci_integration", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4ece9960f4015e40f33440c9629bc51e85bf3ca"} +{"id": "train_09022", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b7e595ae2927e7f6af7e88ca31aa2cb7d1cfc25"} +{"id": "train_09023", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3a8fab7bca6eeaa6a729c709471146e7592a29b"} +{"id": "train_09024", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b22b4b49fe844f469740b8a414868b0e395b56d"} +{"id": "train_09025", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1e0d6e0ed0749c7fe9bdb98d0cea6c08a81816b"} +{"id": "train_09026", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3f579af892df92808160ee8232f7b1592946e54"} +{"id": "train_09027", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9767be1bcbc46eb29fbb7ac05fd7b68261dc6c87"} +{"id": "train_09028", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71117332369626c760d9a99d33508ea2f3c20309"} +{"id": "train_09029", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3c5d0603bcd2cf1a93edbbca40983e6373e46c0"} +{"id": "train_09030", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e37b3de3e656ef70a16299541ba9bb56d28252af"} +{"id": "train_09031", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "tooling", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05508d74d356669158a310d970800e9558f6b59b"} +{"id": "train_09032", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec5d3967c2057f14425e7258dfc3929d976cf5bd"} +{"id": "train_09033", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "repo_scale_reasoning", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5de0cc8ce38da6acd81bffde47032dbf29287000"} +{"id": "train_09034", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "auditability", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac7c94740920cd2959f435945b7f13ff1d0c804a"} +{"id": "train_09035", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0a1ed613cc474eb3c6205201bd4efef1ce9e12f"} +{"id": "train_09036", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "913d82e6bacb643ea126cc836414c6d8a123321c"} +{"id": "train_09037", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "auditability", "reproducibility", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d9dc9ee418c302de18fce55eb33852975657397"} +{"id": "train_09038", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "358cb5ba6e2a76af728ec90ccf57dd36b6693328"} +{"id": "train_09039", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95da4c8022bafeeeb7ad288c9e4f2e9115e724ed"} +{"id": "train_09040", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "documentation", "tooling", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "148dc2be58d7c2e4ca181c230cbe97688282a621"} +{"id": "train_09041", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e02c712e26524ffc73fab277f13c283d3e00664d"} +{"id": "train_09042", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "ci_integration", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3a715e3150c1709eaaa20d73cd825a468e8f9d7"} +{"id": "train_09043", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f09a50799d0f1e65a6b357652c6b34b543c4628"} +{"id": "train_09044", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b025663d8e832fa6e7363f6d1ddd4f7f424e25d"} +{"id": "train_09045", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbc94d6388ef42c72246c633191af4bbaf577216"} +{"id": "train_09046", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "ci_integration", "security_gates", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b7f7bc42c28a4b8cb0e343ba6274139bda8d799"} +{"id": "train_09047", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "tooling", "documentation", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bb64c087efcfbc157530a1fa93eaa3124406eb3"} +{"id": "train_09048", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "security_gates", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94a4c8b8ad307bc4f0ca31e1004284011444dfc3"} +{"id": "train_09049", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4edf16aa48c1449aac30b6d4200ed227b538a6de"} +{"id": "train_09050", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12bd84286cc50f6ae173f0987decfc8299ee9e23"} +{"id": "train_09051", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "273810dda51ab2122b3bc7b12c8835c0066c22a0"} +{"id": "train_09052", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b21873863e0e52a0cb3b3d0f7a2b9ec0fe10de90"} +{"id": "train_09053", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb76be466d5103bb5e274d1517d9e4afad99a642"} +{"id": "train_09054", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5aa1ab3e43f3cdf2c0eb5766adcd306e95ddf20f"} +{"id": "train_09055", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c21c03a7e3373588d062ad7e251131af0a1b77a1"} +{"id": "train_09056", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b7c0916aa3dfb7382c410b2bcd5cd815254ef43"} +{"id": "train_09057", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "auditability", "ci_integration", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3993a734a39be74ba0f600c7601aed0dcc3e432"} +{"id": "train_09058", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "governance", "security_gates", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "acef2d4f6a522a6353fe5fe7c7f5523f80796541"} +{"id": "train_09059", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tests_are_truth", "reproducibility", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccb2257b26d84b38866ed0bbd1a2819b853b55f4"} +{"id": "train_09060", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97b3b25ae9e981c12f21e9c37cf2af9a15c52788"} +{"id": "train_09061", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27d328cdfba109b1eeb8efbf4dc565b177bd0fc6"} +{"id": "train_09062", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "security_gates", "tooling", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80008ae3ee0e34f176c438e4f90f7ce6cbae003e"} +{"id": "train_09063", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54a87ac9719b6cc4a17a6961aecacf6b05b28cc7"} +{"id": "train_09064", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af5a3f367ab8f32632d3ef367551a6bbfa8625b2"} +{"id": "train_09065", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4ba27494e2d6b53be2ac5c9bd0b7d532091ca2a"} +{"id": "train_09066", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16272ec565096873c55fb3bfa0f2a78cc4d34ccb"} +{"id": "train_09067", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "cost_latency_tradeoffs", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c3823c8228dc2395ec6a57f2cd570f00f5500c6"} +{"id": "train_09068", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "security_gates", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c07acf296451868c58d692b1d6157a770314e35"} +{"id": "train_09069", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fa2ce7c1e6552cdbf2e793d916dc12e66e44518"} +{"id": "train_09070", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "ci_integration", "documentation", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b62ebdd342918a9d72f7e83b953a7bbd2e1ef22e"} +{"id": "train_09071", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "526c3c1483c97cbd9c6a67c4f37419a6cf611d97"} +{"id": "train_09072", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "documentation", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b451d21804967448864db4551e30ce984fcd9070"} +{"id": "train_09073", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dcd11194239811c92896469931eef0ce503c501"} +{"id": "train_09074", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "governance", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "996b7b00e539edeb30fe1ed521f725939d5ca80f"} +{"id": "train_09075", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0386cccae27f17e584f60e1a27ad94fddf7ad7fc"} +{"id": "train_09076", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b945fef042b9a5770b26f413d1c2f8d3b18bb4b"} +{"id": "train_09077", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4ee6163e373b8c9d857c0eba9eb49912bd83f6c"} +{"id": "train_09078", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24816d0e44a6a4cdb571ab0d5fd13a56c2495903"} +{"id": "train_09079", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "ci_integration", "governance", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc9b760ce9b5d494410944a43474446959b2156f"} +{"id": "train_09080", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17b298e3e38cb538249375fb52b005b5b4d9e26e"} +{"id": "train_09081", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3f4266e804c22d4a2844f6bad8ad679fd44e1e0"} +{"id": "train_09082", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "859195a352b8cf1c0c739006e1da21c39c46f42f"} +{"id": "train_09083", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2672c5a5e34dd2dba0e03148227a0f5cfffa134c"} +{"id": "train_09084", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "auditability", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93d2456f0b762c01511b66aedc6a20a0051b2491"} +{"id": "train_09085", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "191632d59372c077e9d19899d269e7ec4148bb97"} +{"id": "train_09086", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "tooling", "security_gates", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e849179b75021ed34cb0e8d765264e1f173f0ca"} +{"id": "train_09087", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "977712837ed496e49ce7b75ca41f787b73407c67"} +{"id": "train_09088", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "auditability", "security_gates", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45b5a58d27d543a3d77bff4184afad49f09f5f66"} +{"id": "train_09089", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "tooling", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3875f0f0aa729e265dbe052eaf28a236fa938ed"} +{"id": "train_09090", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62750cfaefab5f99033dd1bb38cee07855f4cf56"} +{"id": "train_09091", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3df1834f616ccd5911db23df0afc14a6f3b6b79"} +{"id": "train_09092", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a1db28b9be36c662ccd36133ecedba893f7d346"} +{"id": "train_09093", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "governance", "tooling", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f0b486ce2ff684874510623818d54dd1feb75cf"} +{"id": "train_09094", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c449d71a4bee062e7787ef5455708f65473ecef"} +{"id": "train_09095", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fbb2f13a0aa8b2e5303e0b3a8f2133d7a91178e5"} +{"id": "train_09096", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67bb1602766f14fc7668491da1ca5bafd22905f0"} +{"id": "train_09097", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "reproducibility", "governance", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e08f6538fe459adf36f00b7a59a6922213d9686"} +{"id": "train_09098", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "reproducibility", "documentation", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a4d8531fc9ce51cbb7cc61f8f3b20098d1844e9"} +{"id": "train_09099", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a9aa53c7fc57fcaae65bbb8a5fabeaa2c4fb331"} +{"id": "train_09100", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "ci_integration", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f6e72bb0f2bf72c7f98b65121af341555003ed3"} +{"id": "train_09101", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "231e878e8e14bd81834f5a2a65abf969d2b268dd"} +{"id": "train_09102", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dbd36e2f7336989befe83203722b7ea0a8aff59"} +{"id": "train_09103", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7250651ac0676079cb424d8e9cd19535bd1b2fc2"} +{"id": "train_09104", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "governance", "documentation", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aed8eb0cfaa56de2938213e3cc554552c8666796"} +{"id": "train_09105", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd692d6294575f3a11aa46bf6adf503d1d3a1480"} +{"id": "train_09106", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "990635771aced57a4c832aceabfc74589bce7306"} +{"id": "train_09107", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a76fdc819c67406c69017775a2e039a843911660"} +{"id": "train_09108", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "security_gates", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87e4c660c548af5e3509c130e9a60159468e06fe"} +{"id": "train_09109", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bea488ce503408c5723cbb2e00cc07e8d8789cb2"} +{"id": "train_09110", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a881723bdc47aa242a457e8178417968829525c"} +{"id": "train_09111", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27e1bb0f2a9fe22df6ad6f499c0d3e173cda4ec0"} +{"id": "train_09112", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tests_are_truth", "documentation", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be18ece06260afcceea98c689d09e82fd8aab25f"} +{"id": "train_09113", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6caa921d04a9a1a712d8e1a3a68eb7abf405c89c"} +{"id": "train_09114", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "documentation", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e95c3d7bca58326773cfdb5e21ab4cea1b856ce5"} +{"id": "train_09115", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bc9ed12ff71c364809defe21a623a16abef67c9"} +{"id": "train_09116", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "474a061b8e10abdc5be295786429e0c273eed04e"} +{"id": "train_09117", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b93541b9e97a019e6887ba18d7094d5d1fe61ca4"} +{"id": "train_09118", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af0196acc10a5533c518c94d8a4fa8fb55b4c201"} +{"id": "train_09119", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "security_gates", "documentation", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43fcfd556d5e00182c71b2736a81f26e875f0681"} +{"id": "train_09120", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38ea56ab28bf11fe897511252f8bd7dd423edc17"} +{"id": "train_09121", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d16f068cd5f44e0520eb3bdc35e8a686bc37a6f0"} +{"id": "train_09122", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "documentation", "tooling", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c2b759112b578d0aae3c62418f3b2dcdee3373d"} +{"id": "train_09123", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "tooling", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efec2bc38ed2825ab033b00e306e5e60abb76315"} +{"id": "train_09124", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4c08fcfed1d42236b874b767f4dcc73ec369154"} +{"id": "train_09125", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7d2d03d1b3af80158ffb533ff2f784a33d8fe9b"} +{"id": "train_09126", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf79aa361c239ab94768f67979d9a6c92d1f2a92"} +{"id": "train_09127", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e63984ee5f3f19fb869d7475d446de9f1a7fd85f"} +{"id": "train_09128", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93f189aac43befb83356615ed1a4a5957fefd67f"} +{"id": "train_09129", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc30d6c375eb1a84459f338714e110b18bc8aaa3"} +{"id": "train_09130", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3af6651ada8a9c9c069376490083ed2de7553cbf"} +{"id": "train_09131", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3b84390f20fda74dff4c8954705e31e32e364f9"} +{"id": "train_09132", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1adcbdc0422b234349a42cba298085c52a386b9e"} +{"id": "train_09133", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tooling", "governance", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "855fd810ffe59c02b166509d0e2abd6d5aad6062"} +{"id": "train_09134", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d457ff1107f2bdb27888fa1b603d8db837b571f4"} +{"id": "train_09135", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "tooling", "documentation", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ddf4968006b4f99972691440f615f3333d9c723"} +{"id": "train_09136", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94c97a6903320e2cf4ede1f417ec265ba5ae38bf"} +{"id": "train_09137", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8657d025ff2fe6778a2e5c4f86f5e6912240bb9b"} +{"id": "train_09138", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca5e5447cbbdf790849d38141c4ff59f07e5756c"} +{"id": "train_09139", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "188c24936639ee4659e6ad9420cf2aba638a28ef"} +{"id": "train_09140", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0970c907721c951b5ad61275587c557069bec019"} +{"id": "train_09141", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b742d0243c92c8a3005dd83e92740a357544a10"} +{"id": "train_09142", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe9523f8281f043fe9c56172ab706ef3850ed93a"} +{"id": "train_09143", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "187c40531500ce24e220626003d62a774bb1a339"} +{"id": "train_09144", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a9dd9940c14063f603b3875313ce534a5403817"} +{"id": "train_09145", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "153e5b4e4b53b0d99082d072edee2c02ae3cbebc"} +{"id": "train_09146", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "security_gates", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8702533e65c9e003e67116834e63bbde136fe94d"} +{"id": "train_09147", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24a29052d16b2971b10a19c139d248ad8101035b"} +{"id": "train_09148", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61c7cb644a16228bed60784ce424abebd1e87b15"} +{"id": "train_09149", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "388d2676ca228b3ae4d8db42fff14746b7291c0d"} +{"id": "train_09150", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "security_gates", "auditability", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2b3db1796d1b8aec8d080a1e4e32fedffec4a25"} +{"id": "train_09151", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfdd54816012e1c1f90c27ea3d217fdd2b707ec9"} +{"id": "train_09152", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f0a5d117518ef035fe7f785b8e6b254bf572e98"} +{"id": "train_09153", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ec542ae797a3328e0f8c674ec59913a8bb44090"} +{"id": "train_09154", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "685451bd4d9e1eb1f44da9a0dedfa130b3f46c32"} +{"id": "train_09155", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "security_gates", "documentation", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab7ec0c4693dca91677172b877054c9cff598e2e"} +{"id": "train_09156", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e939a7bb0b52f20dc5e07093495a493a7c480cca"} +{"id": "train_09157", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22db0466a047c9d5e2fc31931c25d4816a439fb8"} +{"id": "train_09158", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea48c8aa30214505217c79ce2370f16a2d017ebf"} +{"id": "train_09159", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "security_gates", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "341184d4d020c552f1860e762339df6399a41d1a"} +{"id": "train_09160", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e97404e0e4d97177d29e602a2df4fec0a4692661"} +{"id": "train_09161", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab912c4215bdab613bd01cc34f61601546ce56aa"} +{"id": "train_09162", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "012d5d5ed59b34db9c3eb923e638b10ae8468ccd"} +{"id": "train_09163", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7701138d5a33d5a498fa9b90c0bc0e8fd2978a66"} +{"id": "train_09164", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "evaluation_metrics", "documentation", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09cb6624076032e5c275cfc100fb87ef9a313482"} +{"id": "train_09165", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8bcdedd9d202d4e02525aa497d45fc2f9d80b57"} +{"id": "train_09166", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47bd91644d9e19bb33037a3fa9da40e8a2748502"} +{"id": "train_09167", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5799587b1ce38b71417f9833a5de8c633ab38d4d"} +{"id": "train_09168", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1424c4013100c480e981172a40cb52b6f7376f79"} +{"id": "train_09169", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a4a5048ca10e7709d6df5853b33ca894d2130cc"} +{"id": "train_09170", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0112017e6ae9b67a16cce2056db9effb99ae28a1"} +{"id": "train_09171", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40174e19687055e4c49c9afbdc75b726d1c69e71"} +{"id": "train_09172", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74e834bc3b93e47a9ce7019be6653eea991d643e"} +{"id": "train_09173", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tests_are_truth", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7c722963563fab13abe34f5b06844451af55cb0"} +{"id": "train_09174", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92e8e4a5ccc97eeae15e4da8a19259b7c90ab552"} +{"id": "train_09175", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e2673464801bdc44dd9a25070942471e5c208f5"} +{"id": "train_09176", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bded7d763e429772704bf0679b35f823ca861256"} +{"id": "train_09177", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "167b25b35c6bc9229f8c0051d3a50c0548907bf6"} +{"id": "train_09178", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d1cfd25e1fc65708223b952a0b43da92859aafd"} +{"id": "train_09179", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "reproducibility", "documentation", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cdc3a99a2f0c64264016ba26c0649ae14332597"} +{"id": "train_09180", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d49762cbad1c4fc358cebb75ef9b0bb6b96f424"} +{"id": "train_09181", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c417973b73ed4f2c7368cd4c9ed89719c5a18e1a"} +{"id": "train_09182", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "governance", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e37dd6059f3211b3107f391297afdc2164b08b2d"} +{"id": "train_09183", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "governance", "security_gates", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e487d0618e89ec583486a6f48c8f59734f5defef"} +{"id": "train_09184", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "security_gates", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b43b2833f2668ed5bdb9ed9c9c71dd15fae53efa"} +{"id": "train_09185", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "088069f8873c2edc9fc14ef5d0b321e9e367b917"} +{"id": "train_09186", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc41c78c3eb5eaefbc017f07b73ce9573fff05a7"} +{"id": "train_09187", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5914219d38a77dc96c9cb0b98c9df8a30829fae"} +{"id": "train_09188", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "governance", "tooling", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d341a3a923cc2d0a2ed74a35d8a4a6fced85ae47"} +{"id": "train_09189", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9f0b6172b744634768151c92c1bd76f65e30328"} +{"id": "train_09190", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "461df607e26b8404a531eb9a67314cd1ab6b2b76"} +{"id": "train_09191", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ed3d244c5e91f249137ecb5c3522f2bdd86f9cf"} +{"id": "train_09192", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "security_gates", "ci_integration", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7583497c97360104f2198f228833cc36e221dac2"} +{"id": "train_09193", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1498765568d5f6969778c72e9330f59be161a97"} +{"id": "train_09194", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b20ed7e02913654823ddfce90bf117390e8c6af6"} +{"id": "train_09195", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9aed69706d9cb279f1d22267731e245b98ffc0b2"} +{"id": "train_09196", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc0afeb7a5647ce4a8b0a32108788eefe2bfad55"} +{"id": "train_09197", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "reproducibility", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "759621acdbf7753bbd62e18f8c6261dad1d22ddb"} +{"id": "train_09198", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c7c1cb7e2ec5ab0fb8d023e4dd64215fec46125"} +{"id": "train_09199", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "governance", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a78eb017504a5016b19d707d5cc9c474f9b08af"} +{"id": "train_09200", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "auditability", "security_gates", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2505951791972a711415e22f95407951e9d76d0"} +{"id": "train_09201", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96da1556bdc4f2a234c4a1c2d0beb5595c71112c"} +{"id": "train_09202", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c717552fa3ed87ffacf442237c904afb5e40ba75"} +{"id": "train_09203", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97b3b25ae9e981c12f21e9c37cf2af9a15c52788"} +{"id": "train_09204", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "documentation", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bc9209a8a922f7ee86145d85670b73f1d070100"} +{"id": "train_09205", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21247bf6ec5b436158de9047059e8f59fb2c5d80"} +{"id": "train_09206", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "security_gates", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34941ab4efdb8b5fe48d42dfc74e0402457f6ae3"} +{"id": "train_09207", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "security_gates", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc002777883df6f63bd1d1efe50aa0348c651427"} +{"id": "train_09208", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf18553c651da942990115602e6149fde521136e"} +{"id": "train_09209", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24c9afba8f17a374ad9364758189b6fd7ec7ae9e"} +{"id": "train_09210", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e1a10dabd06958be9ee99bf1d92fd5a1d8eaa9d"} +{"id": "train_09211", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0004acf31fb01a12ecd583dd4b4c7343c895d495"} +{"id": "train_09212", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37996667133047679dd6f8df43bd45ce1e6f8e8e"} +{"id": "train_09213", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "ci_integration", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52bd0ba10ae7b4bc179b42b6c8359d141dab2b4a"} +{"id": "train_09214", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "056553ed340e91da03ae1a68ac37e1532ff46ec4"} +{"id": "train_09215", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98839e2634cada7c50404f3ef0206fdb128acc01"} +{"id": "train_09216", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tests_are_truth", "governance", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2657afa036d5ed5ce553194a90582e24c8f5c81f"} +{"id": "train_09217", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ebdcfee8b44cb0af97aefe46cd172e6e1b3ee85"} +{"id": "train_09218", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e666cc1dafe2d9fea82d0b0c799109ee12d67c15"} +{"id": "train_09219", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "evaluation_metrics", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c5d4457b5a5e0f71b2b8b494936bb420381e1eb"} +{"id": "train_09220", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc7551574aca8172ae21a23e2cfee1fb8976bce4"} +{"id": "train_09221", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc871a73baab89eac9a468d87f671832351f1b6a"} +{"id": "train_09222", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be618f2e6bc8327cd4ee13a3ddc5bc91cae46333"} +{"id": "train_09223", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "documentation", "governance", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc682a53d95dc32aee5ce70b017aad3fe88a7203"} +{"id": "train_09224", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c63cba48739fbac651ba47f59dad4536680b3f55"} +{"id": "train_09225", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4672362ba7d670926184ea16ac10587a624c20e"} +{"id": "train_09226", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef985048ee27a820a8a2856c0e9db98b9ded104"} +{"id": "train_09227", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87e7642838774dfd635cb91ebdb3fabd23c550a7"} +{"id": "train_09228", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "ci_integration", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ac14f013f59755f329996b10f1159a6782d35ce"} +{"id": "train_09229", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tests_are_truth", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2fff2356287a1d9dae7f8e8df91bdfe60d37322"} +{"id": "train_09230", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f3d75f90942c7195f3af59b28c8213eb65a52af"} +{"id": "train_09231", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a073edf14a66667549956294ff4b3f5dee72c18b"} +{"id": "train_09232", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8be7d3b3a37f81c3784025f192a753631920a79f"} +{"id": "train_09233", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "364bbbba4e829f4917859b21b1c408fcb37fe7e6"} +{"id": "train_09234", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7389093a43570315b1b73f63c5dcadb1803579c3"} +{"id": "train_09235", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f520e364f5d68ac498fabe6552df1954381faeb"} +{"id": "train_09236", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "auditability", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb97228f9e12b2b7a31e9e918141b7ad3828288e"} +{"id": "train_09237", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "documentation", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45d6e0b3ee4641465c2b95a7645f3aa239447aca"} +{"id": "train_09238", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e312535c26f2780917f94d6420b435e2e06eba7"} +{"id": "train_09239", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "repo_scale_reasoning", "governance", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "adaacdb406c7b90253e4d477ebdd7eceac219de3"} +{"id": "train_09240", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7823b34a9d3b49f2120ba4ee7ad570b23f85f0cc"} +{"id": "train_09241", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07b71f88ea4ca89d74fbbd7389703b323f7847be"} +{"id": "train_09242", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8d2bcf123a78e1ca7737c61e3b18d43003c68ce"} +{"id": "train_09243", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6c8036cc389404d23adfa7dbd8edda1b81f6884"} +{"id": "train_09244", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d895ee385a3c6c1ec673a75632c00891b99ef79"} +{"id": "train_09245", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94124218079618e89bcf13b545a0c6568cb8fed5"} +{"id": "train_09246", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc00f574c5d10aab2be2bd4461d8bf0d144b73c1"} +{"id": "train_09247", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "auditability", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bb3586a2d8efab3ab6dc1c452696d440240a980"} +{"id": "train_09248", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "security_gates", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6552e60735050a2db9005e4337220a5fc05cd814"} +{"id": "train_09249", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c5483d89b61021f96bc0a67cb8a9f16babef0d4"} +{"id": "train_09250", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cb12ccfcec64edc2deae4ac26fd7bada0728751"} +{"id": "train_09251", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8da69c0a4048771139e1ed9de582af0d623ff43"} +{"id": "train_09252", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f35d5e99fa3e6dee811487a7c3949f43e961516"} +{"id": "train_09253", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "documentation", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "964251bcb2ec9dd3d931752262d0cb0264abd797"} +{"id": "train_09254", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c86a6c2e19bbc86e9711fb63593394cf29eb27c"} +{"id": "train_09255", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tests_are_truth", "governance", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50aac3723e8088e5a0d5265eb0489ff0084aac20"} +{"id": "train_09256", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5eb51cf9aaa361d23cb3561c89fe449713e05d8a"} +{"id": "train_09257", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "documentation", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "244994289d18e2ffafbf8b07e15da9884ea85d4e"} +{"id": "train_09258", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "documentation", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fe84af7ea9439cddb6bbb89ce5ceca49c84ee84"} +{"id": "train_09259", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "022ab8a5f14dccbeb2685bd6cbb03e6503b33566"} +{"id": "train_09260", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "tooling", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d01672dd74c62b90ebd1fff46fe25059f615c72"} +{"id": "train_09261", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d022aa125557194f8b9d7f4a378903a87f1d4ef"} +{"id": "train_09262", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1137d3adce98b56bee53c9205d61849703415b1e"} +{"id": "train_09263", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "auditability", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "320b5b0455526ae493e9419f3b637e8b290385db"} +{"id": "train_09264", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77c6ea74bc50027909d9b91664f6e67d500c54f2"} +{"id": "train_09265", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06133f698008984cbb193df66256a0e434ec014d"} +{"id": "train_09266", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1dc1ebc017da68efe6342af346b7b52d2bf23fca"} +{"id": "train_09267", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95e277a13790952585f626297d6e2d054a131d2a"} +{"id": "train_09268", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea76e05f7089ebe9db4227b73a0c04b001ad0792"} +{"id": "train_09269", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "auditability", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f897d48f7780574545bef0eba759604564b9386"} +{"id": "train_09270", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eee0353682455c824cab04b8e420414c84113b13"} +{"id": "train_09271", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4bc0dc9470346a4e52b2f0ae4498e9cfc1f6b8f"} +{"id": "train_09272", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd3e93ab2f48024c67631d4876de6e654c6b2e42"} +{"id": "train_09273", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bf82ed6b38a2cbca1d4e506f48f9ce6bc21b8a7"} +{"id": "train_09274", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20e7f57364324a0cf3e50c7b55b4a4881271d050"} +{"id": "train_09275", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffb5af248f6d71fdc182b1995cd227b257e4599b"} +{"id": "train_09276", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b2a6142660ea4660fe89681e61bc77c070b0993"} +{"id": "train_09277", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf64c120fef56e278e0e816703be07b4e9f54d6b"} +{"id": "train_09278", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b50574c26689ca0e02bc26cf2b5f94af3cdd3c60"} +{"id": "train_09279", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "governance", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc0841be00889c403e656eab4a5d78ba6639b0a4"} +{"id": "train_09280", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fa0494d304ea6aebdecdd188b2bae3ad6d31b94"} +{"id": "train_09281", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "governance", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aaf223a5192fec848b33ecb2e04c26d1bc10eba9"} +{"id": "train_09282", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "governance", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34416546579f1149b724e5879670c2363ee988d7"} +{"id": "train_09283", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c98f225ebe003cc7700d8c0680980ebcc141e0d8"} +{"id": "train_09284", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bf33ff08016870f481522867666ebdd3d5cfd2c"} +{"id": "train_09285", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16c68b0f72e634ab7fe3fb5a78da68d365befd3a"} +{"id": "train_09286", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f408a23d49edd20fe46dfd3db1294c7cb9ef2b5"} +{"id": "train_09287", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "governance", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c5ccba11f2696a1d1b7b96eb9832bb7045c7bf0"} +{"id": "train_09288", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tests_are_truth", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "863344129d739dab97e78a8fc2615c7cd8f5149d"} +{"id": "train_09289", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f26f66928460ea15d250577189cf8cbb5d9258d"} +{"id": "train_09290", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "evaluation_metrics", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d18408230c07b78df116c197869959e443906578"} +{"id": "train_09291", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c13f4c42da8a0993d08466908d1889f2e1730104"} +{"id": "train_09292", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "reproducibility", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f36961a322e49abfa9c35d3a85168fcb0cf2eaea"} +{"id": "train_09293", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdc91c16369ad1af94b28adcd23e36dd3f625fde"} +{"id": "train_09294", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "documentation", "auditability", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da49f2ef73a859a84b182d697d2a32d185786253"} +{"id": "train_09295", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cc92dedf1f08512af6d4d27f027921f071a52f2"} +{"id": "train_09296", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cb663132613624453fb5c5ac57199b9f9ddb292"} +{"id": "train_09297", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "ci_integration", "governance", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56cb4d94358de1733c2373661157fb2a0cd3896b"} +{"id": "train_09298", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "reproducibility", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca94b50467dcc31e257bbf61764b7ec3e6a9e916"} +{"id": "train_09299", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "320b5b0455526ae493e9419f3b637e8b290385db"} +{"id": "train_09300", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86d13ef34d55cf2d9e1c0420ca7718dbfa1b0b58"} +{"id": "train_09301", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd6300f4313b4713c4e975a56133b70ef13e1a40"} +{"id": "train_09302", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9285229c55202fd7e89a2cbf6bb35e847a9521b2"} +{"id": "train_09303", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "evaluation_metrics", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec4239d7dd33a08460bae772fa6e8fdd50d1125e"} +{"id": "train_09304", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab703aacea3838ffd19bb95ed27e78dfcfd75f3e"} +{"id": "train_09305", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb96315bc481857046a1d69f1252576cb586bef6"} +{"id": "train_09306", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c341e76d2d849ffcd9cafd382ac2dfabca66f9e"} +{"id": "train_09307", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c592d178d54f58e2d00381a7385094df7e101209"} +{"id": "train_09308", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "documentation", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "329d796af2884446ed8dc8744092bd5e0099730c"} +{"id": "train_09309", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96afed11001035b188f840f98cdc74f508fa6caf"} +{"id": "train_09310", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "256b0d37e81412882880d5424e22ab2dbe38788a"} +{"id": "train_09311", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "governance", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c542954b4d9e37873d17e178ab9e98a83f4e016"} +{"id": "train_09312", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05c858eef2dbddd0765ec8d596611489da6db91b"} +{"id": "train_09313", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cffd4cf5f5c4895ca3ea6854f643480716a5930"} +{"id": "train_09314", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76a6ffc95242d9f106554cc84458537980a4c5bc"} +{"id": "train_09315", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f55c1219d91baf8b1282fb27252cfc38d552783b"} +{"id": "train_09316", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7e957bc76cf45f231fe4246e3d5622ebff12ed7"} +{"id": "train_09317", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f80d7373d273fbab80b7aace57b44d2d5056bec"} +{"id": "train_09318", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "reproducibility", "security_gates", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "864c8afc44cc91ed9979c171a3d3b2d29a24dd75"} +{"id": "train_09319", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d61b892477b2b5813e84153073a95f8bc09a9510"} +{"id": "train_09320", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dae18790b7006890d81afd26cac4c0d1dd85e5c"} +{"id": "train_09321", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "security_gates", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46acb660a144f8092a17ea0aa0f8d63bc75e40d8"} +{"id": "train_09322", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "ci_integration", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9658dfca0b17186e3c88481254fcda10fd50a9c4"} +{"id": "train_09323", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c5bf6a351139aa6aa673fc582159a972a4eced0"} +{"id": "train_09324", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96cb02b9b498942b56d34e8be5077f376976bec0"} +{"id": "train_09325", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "ci_integration", "tests_are_truth", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a4caea74258787176aa9f6373d993e94b5fc329"} +{"id": "train_09326", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfde98547c62e897353db2ce00e037edadab8238"} +{"id": "train_09327", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ef8bc00d380382f8eda3973437c63196a420bba"} +{"id": "train_09328", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca8da8419db3c11149e6f97250519153584ce267"} +{"id": "train_09329", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "documentation", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3df60378e76127a9ecd773eecc575d50bfaea328"} +{"id": "train_09330", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b013c9b886c5fd90f4a9efd1b0f7ad88fcc51de3"} +{"id": "train_09331", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4b58078bdbb01be2de220d9b255377d7ac58ae1"} +{"id": "train_09332", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4f229db37e1afb65304e191e73068599ba3dacd"} +{"id": "train_09333", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0fd571d880232278e6c6bfdcc4371926420bc7e"} +{"id": "train_09334", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a209398280ddc15f8cd997f45911fbd1eee3d03"} +{"id": "train_09335", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "731a295a8a178fe203abb59800e92c3bdaa43174"} +{"id": "train_09336", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "faf3a099522c3c907843c5c95dbaaa0e33b99f75"} +{"id": "train_09337", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73bb0b79897bbdca9d09a32eb38b40f35d52ad89"} +{"id": "train_09338", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8045e5dbc54084a77c6b22a83f4a898153183931"} +{"id": "train_09339", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "reproducibility", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c838a4bef742d5221001f9e8faf0f23af0b39b5"} +{"id": "train_09340", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7e37f1a020f3539e2759932f9c6507fb4c57ad8"} +{"id": "train_09341", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "documentation", "reproducibility", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cf546a5b0dab84c0c31f809fe1fcf48e28aa63d"} +{"id": "train_09342", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86182081707b47e70fcac470ea12d3264f450c82"} +{"id": "train_09343", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tooling", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f24f52ebd8465af9ab0427e174d5f16c1ef91f72"} +{"id": "train_09344", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dccce11774c8740c0c2445ee7f23b41ea8f6fe7"} +{"id": "train_09345", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "reproducibility", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "607e76087610319a619ee82d30ce9d313ec19ed1"} +{"id": "train_09346", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02522b3031268219b99c71a4e2f0dcbeb923c54f"} +{"id": "train_09347", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "auditability", "security_gates", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e75face8f95eeebd2a17815399c7cf8177b1bf7e"} +{"id": "train_09348", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba72e47d5f99ba421dc4a34547c9351f69e873ac"} +{"id": "train_09349", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "ci_integration", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b732ada121f25c04ec7c3f7cf3d336159823a26"} +{"id": "train_09350", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "documentation", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44ae5cb5e3c4669f0a095a46913905e9a383571e"} +{"id": "train_09351", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9a79d0791568ca46f8c2899bd0b8c1f6f798902"} +{"id": "train_09352", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc91147bfcc886e6845c7282b45e475a7862a10d"} +{"id": "train_09353", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fae6758a4cea977c8d3dc2de4146ab18b0a0b444"} +{"id": "train_09354", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "documentation", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7cb61e461cc0b6977049e2c011cb4483c82040a"} +{"id": "train_09355", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "governance", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "500fb53c6ed71b33cd8a868ad94e5ada527684a5"} +{"id": "train_09356", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c1c107ff432b042b1ba96c77b8f9f40e54b6a1d"} +{"id": "train_09357", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "governance", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4235c0a47708143c2dcc01d24e29796e1772caa6"} +{"id": "train_09358", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69a4e68b1202f1950042bad8a596db7a8219896d"} +{"id": "train_09359", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "security_gates", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce0b8f38d460f94ca5246a6f24e41b0b1c5af70d"} +{"id": "train_09360", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de3220e90f0bd956dfcc6d80cbc46b16ea9fad4f"} +{"id": "train_09361", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66c4c3be20f7c52f5c58fbfa31507ddb02cc1d4f"} +{"id": "train_09362", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "tooling", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c508cc7dbee7a0a6739d2c97d0e39159f5061fbb"} +{"id": "train_09363", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c5d91a3b6ec14ae2a776ee594a5ade492dc283a"} +{"id": "train_09364", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f23ecb5e6357c5f1ea81f216f617dea2c2084701"} +{"id": "train_09365", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "auditability", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3334d35e5fa7cfef95b2e757e1331d70d235f62"} +{"id": "train_09366", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8587b24e6c2f45855c38f0718960a31ff6a435c9"} +{"id": "train_09367", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "reproducibility", "documentation", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "670ae1c139a6b4ef4b5534227faf48647385d50d"} +{"id": "train_09368", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b568695557495741a4a6aa5dda9d3bbb4ff52e9a"} +{"id": "train_09369", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tests_are_truth", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d332fe69c3d7f37cae2d577b1eec002279891b7"} +{"id": "train_09370", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "708a70dc5e3c73adc45a5d31b7bd95dd9189bcf2"} +{"id": "train_09371", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcd851eb168fd2047dba9e3cd7c1905259b3e969"} +{"id": "train_09372", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75074d91aff73d217caaf3ea1c32cf9b9cc075ea"} +{"id": "train_09373", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "ci_integration", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf8a50e9b7571398c77f785e707ef1b71f834dbe"} +{"id": "train_09374", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb312fe3fe54853fe3a04056a9aef5779720d571"} +{"id": "train_09375", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "documentation", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5859e6a49df1135e92573ab00d3e480f5cec1a8"} +{"id": "train_09376", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8099432c612e709225eac43f65871f4d3c6b4170"} +{"id": "train_09377", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69ec6417b6963909eeb29fbe1a21a02b7f4d39af"} +{"id": "train_09378", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7de2f80ded34c23528a585f09311dfb040a5075f"} +{"id": "train_09379", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0ee28554b6237f11c4a30eadda5691955f58c36"} +{"id": "train_09380", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "governance", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "238839cee8c69e3eb761656987d6ef154e2eb73e"} +{"id": "train_09381", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c82c9d1889e38e0689a9b4bcca799d329a761e2"} +{"id": "train_09382", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4eab1281cb5540c3a86cd54193092f2b43cf326"} +{"id": "train_09383", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37ca5d1c68ef6bbf6f412da295712fcaff1a7fe8"} +{"id": "train_09384", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84d59c1a35587f434dea52ed2b29a60ea6531cea"} +{"id": "train_09385", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dab2960daa8912ec0bc3442a18c46ea12d5542d9"} +{"id": "train_09386", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "auditability", "governance", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "000caf1b59c83c39e603d21edb07487f1883c9a4"} +{"id": "train_09387", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "379eb6b2ba9932c9c10a6da5097bdbb239dbb9d7"} +{"id": "train_09388", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c11f990480e4dd61e66c980f67f11ba27eca552"} +{"id": "train_09389", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "reproducibility", "auditability", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88dd3284723e036e0fa4ee9b4211be1a70600250"} +{"id": "train_09390", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84dc69c88b1b0d29131081ba7559b36fee02e3ef"} +{"id": "train_09391", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "reproducibility", "governance", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0054961b55b4d7441c02f9556ee881bfae5a6a0b"} +{"id": "train_09392", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "097fdd55ccf4a444eabe77b9a86cd73bf76c6c92"} +{"id": "train_09393", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "security_gates", "tests_are_truth", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b42ae29603c156b9329451b306f4da5370acd4be"} +{"id": "train_09394", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24eba2181ca53ae193d6babf60c7a42f9c106c1d"} +{"id": "train_09395", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "governance", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5517bdf873716f9dd49e739304ad3db98aa763f4"} +{"id": "train_09396", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfcb6a561297ddf12d3abe177a266b47e816ab3a"} +{"id": "train_09397", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3516c3dde1c121aa350e854518184f011105bb74"} +{"id": "train_09398", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "governance", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56cb4d94358de1733c2373661157fb2a0cd3896b"} +{"id": "train_09399", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7d05e3e82d93f93d5c8eaac5671dcea905aae47"} +{"id": "train_09400", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "271a063f07b36e06f1871800ab22090c12ec4ee5"} +{"id": "train_09401", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c4d673780ac9349498c5848f88056ba700a73c1"} +{"id": "train_09402", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ddc363e4162e93704dc8e08d3aa8a6869cfe7f1"} +{"id": "train_09403", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "governance", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e3e71f24b5d30cc9cc704855c9e90ced253569f"} +{"id": "train_09404", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "auditability", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c16f9d2e15d634cd698d9c41ea7a50ad8a92e190"} +{"id": "train_09405", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "security_gates", "governance", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6622daca6567d2a3dd54ac54ae56000f25334493"} +{"id": "train_09406", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d46d09d96663053e2972a2d97b91ac160c1e189c"} +{"id": "train_09407", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4932a49c7a45c8bf72d90ae171d6aa39673b68df"} +{"id": "train_09408", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "security_gates", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38092352a439926d0dfc45355cb5e7944a3df2c8"} +{"id": "train_09409", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89c072a67e24b89dd9237a6d1669891105299965"} +{"id": "train_09410", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "reproducibility", "governance", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca093fc201fe2e1d59f03b5175ac535e24dcfcd8"} +{"id": "train_09411", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "reproducibility", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aece5cd83cbe25c14c74fd7995ee22522f2fad06"} +{"id": "train_09412", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "evaluation_metrics", "tooling", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8a3f89c15751998dfc8c5dd3a4b053002f70fd3"} +{"id": "train_09413", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "542bd5144fbfb125b86d3123a4951b3715b6ab79"} +{"id": "train_09414", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "196e961a9294cf665d391accd06b23b435c32b01"} +{"id": "train_09415", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "ci_integration", "security_gates", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b45881ae4e64b63e13b86c2cb75b74e45f1795c7"} +{"id": "train_09416", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tests_are_truth", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3773e91c40cf9fc8d61d240f72e48bde73e759b4"} +{"id": "train_09417", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75b728fb9eb9418cb096420b11dca5feed9b0d1a"} +{"id": "train_09418", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "624aec2566fe43938c5953a7768401ea74793858"} +{"id": "train_09419", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df84a920525517e34858a39906ee28f8004216b5"} +{"id": "train_09420", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40b8349662050a3cd7953f6f88b34d1a2d6c0355"} +{"id": "train_09421", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83bca8180894f7bacebbccf2ef7acbadef4d974b"} +{"id": "train_09422", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01b8a29eefd1a1a97138ce94235284e300c53509"} +{"id": "train_09423", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "security_gates", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e07e2653ccc541c124be7b2ceec15fe0ed8e8af"} +{"id": "train_09424", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2d66175c49d98f2491e806b365fd77ff0662b24"} +{"id": "train_09425", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tests_are_truth", "reproducibility", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbaa0b90478f68b80b527e0b60a39e8a0b34fb46"} +{"id": "train_09426", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "auditability", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4c2363e8606165dcf50c10b1c701ae8d6ab2eed"} +{"id": "train_09427", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "security_gates", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fc126fff15285ec48d349477afce6e1b832e167"} +{"id": "train_09428", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6015ae33fa8444b36a75079fcdb73735ad8bc51"} +{"id": "train_09429", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tooling", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b63e3176505ca874536b1d17c380520b4403d20e"} +{"id": "train_09430", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "339ecdbc2ca38fdc9c300b8ce8dbb8fe7feac0c9"} +{"id": "train_09431", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "governance", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "938a628191983e506e7bd075c5ca7b78dcdb11f4"} +{"id": "train_09432", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fbcc2b88d349fcc12f298c9b1717988d74502e3"} +{"id": "train_09433", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "cost_latency_tradeoffs", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4568a6a944678b12a5cf11e1d270ab29b1edeaaa"} +{"id": "train_09434", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53247b8a7fff1b6c0790f2e76d419f7333cceeb0"} +{"id": "train_09435", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "documentation", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b773a20c817c41cb81f3c5fdb1004cc0b7cc8f9"} +{"id": "train_09436", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a04dbc0bb74e8e9f54a0206361c00b1d46a4f65"} +{"id": "train_09437", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "security_gates", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10a9015957c2f3d8ffa71f7806987016c2e4f482"} +{"id": "train_09438", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e7e1104c6c025c026ae43c0eda2634b625641a2"} +{"id": "train_09439", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "security_gates", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b732ada121f25c04ec7c3f7cf3d336159823a26"} +{"id": "train_09440", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "evaluation_metrics", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "535b1f435d1b2109366c548bcdb295feca6dda95"} +{"id": "train_09441", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6180aebc513b02935d45e333da0f508a12f646f2"} +{"id": "train_09442", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48e91fd54557a76e782766945ce6400965e5f833"} +{"id": "train_09443", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "governance", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a9f38f0c31521bcd382f88f57b834fc6e2f9e92"} +{"id": "train_09444", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "tests_are_truth", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b8727937fc2f7abb8533d971ba41ae703dd20f8"} +{"id": "train_09445", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a1337ac8dc0f5b1fd2e129251c49b001f5166d7"} +{"id": "train_09446", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "documentation", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4830c28ad83d0e531b80e84c65db402503101c7"} +{"id": "train_09447", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61d9a6a7f69eaf30ba6b1afd5145bfea6f1d4e8f"} +{"id": "train_09448", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "reproducibility", "auditability", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f0d27cfd96ef7b356d5aeaf8e45a1d7d3933a41"} +{"id": "train_09449", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e93b099a4c244f1d31e8a65c0ce5045d4bb117b6"} +{"id": "train_09450", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tooling", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0b12d2e86d43629ee236be116f16f0afe34aa4a"} +{"id": "train_09451", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83f60acd9fb32fc158de1f68ad70250f69ea2f8c"} +{"id": "train_09452", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6d89f99e561ec5bd1d7462003f09ddd67045ad8"} +{"id": "train_09453", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "ci_integration", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "253a05b716a12460724a5f0d69076cfc7c62989d"} +{"id": "train_09454", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "governance", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d67de2d229fccfcde35a90cbd2a2c2d0ebabf0d"} +{"id": "train_09455", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tooling", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c366165cf313eb3ec64c7908151fb0895bd57275"} +{"id": "train_09456", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be22701009bca377bf00fb01a84b56f8d02a88f0"} +{"id": "train_09457", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a628556eded50f88c9d445f5ab70d7e239143e3"} +{"id": "train_09458", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61ccad6127a8eace58857f3d29fcc8237a1fc2b8"} +{"id": "train_09459", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da14a12e7b4d9c53335c1457c3806c8e1614130c"} +{"id": "train_09460", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "governance", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0541f5b24583090bec8a237a248a1ddf59ea4399"} +{"id": "train_09461", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68f630f8f6cb042281133e33f9a2774fcc9e5c12"} +{"id": "train_09462", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "tooling", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1025f47f5057afddaf850ecebd8b101b813b362"} +{"id": "train_09463", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "896779aa8bc0725402b1ef746f8c4078313dce5f"} +{"id": "train_09464", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cac5ebd5d0795e1e559bba0fac54c744da1d1af4"} +{"id": "train_09465", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad636e77bd282ccc8a233973407450ec46b2e406"} +{"id": "train_09466", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tooling", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2be4d25d34abbcc007e124920f92ba9f5fc910ab"} +{"id": "train_09467", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "auditability", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "357289dedfa6831c79c6292d23e8bae32d210085"} +{"id": "train_09468", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4e8b8e8ae74c7d471121e3f4a9fc8c00fed8fba"} +{"id": "train_09469", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af2b6e38d7ad66803bbafa7d47aed277a272a78a"} +{"id": "train_09470", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84dc69c88b1b0d29131081ba7559b36fee02e3ef"} +{"id": "train_09471", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62b6a5e234c205b287efbd58a80390f4373d7bf9"} +{"id": "train_09472", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a20997bc106c7c38d6f00ced1a2986852d303ade"} +{"id": "train_09473", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5bac0c568f4b9bbecb350a7cb8af6ed32c424b2"} +{"id": "train_09474", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6bef0c3acf737f6d34e63e8da19230db31fd140"} +{"id": "train_09475", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0578bcdbc7a030635cb558447e7a051f01f269b8"} +{"id": "train_09476", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8948e54f30c06b7b88d77e33af440b59adc57c2"} +{"id": "train_09477", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f056bc43efe55d233a65e4aa2b0544bd5c4aec90"} +{"id": "train_09478", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "tooling", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc76bfd73d0798278e1b76b4f593706f6fc64103"} +{"id": "train_09479", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f239373ce1f5c7ece1bbe8535d0c408954f9265"} +{"id": "train_09480", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4206c5102522007c1ea93284dbbbc97524626b16"} +{"id": "train_09481", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a32d77c89b5b2e532434c3a734672399ade09b6b"} +{"id": "train_09482", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db4803ffbc6faea6a0712fa0e841c953d88ff731"} +{"id": "train_09483", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0544610cc00d2c14fee43b7935985133ec9d95e1"} +{"id": "train_09484", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "evaluation_metrics", "governance", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "059b678fed635ee528dba9c7e12c718c09554a47"} +{"id": "train_09485", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db50ad2a22f650cafe615ef3eafaa2475cd11228"} +{"id": "train_09486", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "ci_integration", "documentation", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfc6f4aa4ba246a63b5206e06a1465f51744a6af"} +{"id": "train_09487", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2b185780732c289f517cb3553fbe11f7aa7f5ca"} +{"id": "train_09488", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56890cd948a09dfca13a73d69e6077fb4eb9e8b4"} +{"id": "train_09489", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdc173b23a1b83f706773cfe3a9dfe82d9684ec5"} +{"id": "train_09490", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f26f66928460ea15d250577189cf8cbb5d9258d"} +{"id": "train_09491", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c16f9d2e15d634cd698d9c41ea7a50ad8a92e190"} +{"id": "train_09492", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56bd5ec8d394785fe4413900efeb81309c7a773f"} +{"id": "train_09493", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "documentation", "auditability", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3f614984ec3bf8b56e1b93c53a5bb433eecbd9c"} +{"id": "train_09494", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b464b89e9944f0f157397f195eda20614f07e62"} +{"id": "train_09495", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13f25c4af3d70a5684820c9403007bf98e03410c"} +{"id": "train_09496", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22794a233163b37a3c2e71ee822cff585b6971b1"} +{"id": "train_09497", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "ci_integration", "auditability", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47a54125cb70508d4c06b101b72ca66f1ee4a4b3"} +{"id": "train_09498", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8010fe499f95803b204053552a4e46e51ceace6c"} +{"id": "train_09499", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57c38ec0d62d59d82d625c84f9355b7de20f46d8"} +{"id": "train_09500", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5d11d10c0ed386ef4c8a90628e9d055720bd1ba"} +{"id": "train_09501", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cc55b2e5733b23e35113d0f543f2529516fb5e9"} +{"id": "train_09502", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "reproducibility", "documentation", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5a603856d7f150fe9b26254f87b2e8bd463d751"} +{"id": "train_09503", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eeee30a4107fdd59b87d30308198ca13eabbc339"} +{"id": "train_09504", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "655bbaf7814764a207a4d68d9d1bbc6795a31859"} +{"id": "train_09505", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1b85779bee02c4ffc486c99bb3aff571ac6a946"} +{"id": "train_09506", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "703b7326a8a6f5476eff3825fc0eccbe6279725c"} +{"id": "train_09507", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37a9d0f22d7879b19d020d1c430f9e5b1c46dc0e"} +{"id": "train_09508", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tooling", "documentation", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e23d570cdcbdde1d5cf580fc74e7dd0fe68986a2"} +{"id": "train_09509", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0170cb245c2e86099f7a3757bc141f4ac3a8ea4"} +{"id": "train_09510", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "documentation", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b6463910e3a799b6745dfa626ef4580d8a64df4"} +{"id": "train_09511", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4ef79b5b420626c37f8a020915993a9d3af648c"} +{"id": "train_09512", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b30cd29a83d697669e6019dfd4bf79ab5871731"} +{"id": "train_09513", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af7ce911cd079db4d68cefe4a9f5207366dd6b4f"} +{"id": "train_09514", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "auditability", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fbedffb48c8fbac7b40b795b86a60cde94b7c5b"} +{"id": "train_09515", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bbdd9dd2d43f45d646cdd431b2d24af8eddbf8f"} +{"id": "train_09516", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "documentation", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5860918d116d86dd0acc121c9a16973feee9baee"} +{"id": "train_09517", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0d601de506267d35e2e1a6088ae1d5ff2babb6f"} +{"id": "train_09518", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1794351da31fbbbc3ae356fedc6d9991b6622631"} +{"id": "train_09519", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3125d830fc2079ea4a3ca2e565e863bfbd2d14f9"} +{"id": "train_09520", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0ef9f68c4d9e3af182d86665f8e55fb6e666083"} +{"id": "train_09521", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5e3668e02ea243df5d9214fbe4d782fcd6064f2"} +{"id": "train_09522", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "660d798ec16b0b7c3eff4d04e9458821532884f0"} +{"id": "train_09523", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "auditability", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da10425d620a49696ec7d819a2997591ff1083ff"} +{"id": "train_09524", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tooling", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1de4eb99959a5ee92c6ee32c2388dfe388f62a39"} +{"id": "train_09525", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "security_gates", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9db6ddff0fe041263ad850dfbee2689d8bd59258"} +{"id": "train_09526", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "security_gates", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "284ab27e7f3883f53170f1c6d3d4753a46afca58"} +{"id": "train_09527", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7cce51c0cd9263102dd47f8d749ff0bd280f423"} +{"id": "train_09528", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "evaluation_metrics", "security_gates", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ccc119b3cde2e71cc00feaec48cb91d8d870c2"} +{"id": "train_09529", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tooling", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47f449d69efd9bc6035542d4e8409ae976cbc8a5"} +{"id": "train_09530", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91fca7fcd65862f564358b697d7d5234d0631ae9"} +{"id": "train_09531", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bdb6d7c83ce1770b0608cbd5d3bc3e2b3f55f6d"} +{"id": "train_09532", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "security_gates", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9fb5d3d49de36e0fe85384c135d3eb041c10a37"} +{"id": "train_09533", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01585e00502632f99e20f3637cd8b0bee76f566b"} +{"id": "train_09534", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "security_gates", "tooling", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcdad25a14898069728c1a07b299d9ec1a57de72"} +{"id": "train_09535", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "documentation", "reproducibility", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88cf237660f1d91d3a420d4445a9585d894fc50b"} +{"id": "train_09536", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "documentation", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3ddff665f1d5bcaa3940172518c0e636bd5bc27"} +{"id": "train_09537", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "governance", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f11136efff252704374d1467638a3530f52fffb1"} +{"id": "train_09538", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "evaluation_metrics", "ci_integration", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7379cc947adacf57c0fdabf1c31945d031dfff35"} +{"id": "train_09539", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16ed11c2e575d4920e9ad4011dc04b2afe44817d"} +{"id": "train_09540", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "auditability", "governance", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7606cf7bb103a4abad41ad28d6b6ec1d6b58a3cd"} +{"id": "train_09541", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1bf5b72b8ece6dbb38c13cf171e23c6a212ae4fe"} +{"id": "train_09542", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e072a054694442bbcad0516c3751bfb67841499d"} +{"id": "train_09543", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "governance", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29aa54dfafe4d0da87f525119d3fa0954f8ea5b6"} +{"id": "train_09544", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16ed11c2e575d4920e9ad4011dc04b2afe44817d"} +{"id": "train_09545", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "cost_latency_tradeoffs", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7fb868abfc8bbf4eb5afd9ab12bdcc048dc7aa4"} +{"id": "train_09546", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4bf4127b3fbdd58db970dbcd2c191737212f713"} +{"id": "train_09547", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "353bd4ca3e1bbe9f6f22af846bd16b5487b7ff21"} +{"id": "train_09548", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "governance", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82e2c922152706ec553df092e302ed8ee3c6b578"} +{"id": "train_09549", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "evaluation_metrics", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6cbcb1a60806d0d3d4ca7a9829e38ad04b3c41e"} +{"id": "train_09550", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6638157c33d110eb7241af4d01f8931af5385146"} +{"id": "train_09551", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2c5a8d80215fce9c7b4ac485731799539a561e3"} +{"id": "train_09552", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c013929b2e4e8de67358df326887fdca7f45d4f"} +{"id": "train_09553", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ef41f2ce9da4a95da6f24ea02c94ac6f3178df2"} +{"id": "train_09554", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "security_gates", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90d653779a6f97353aeeb0edc6c294b4424a59b4"} +{"id": "train_09555", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50e25a64e109256307f21529b8c93581e636f603"} +{"id": "train_09556", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "938652c27ed3074b2fcd514ba569152a5ef5b169"} +{"id": "train_09557", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfae7e09df4e92acf2f6c1b6e1a034ad7514fc25"} +{"id": "train_09558", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8cefa485bb2d5f68501c1fc7830bc66a0f88d628"} +{"id": "train_09559", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f0778e8793622f3e565cad2a7caf1770a1fef7d"} +{"id": "train_09560", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b304abf5822694d96f454fd48d7fc490e64ec98"} +{"id": "train_09561", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "auditability", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e83bc591be444b6eaac14bcf44fb00423aff461"} +{"id": "train_09562", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "reproducibility", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db190b7a948bc71442d4d17d77c95a9f3d1ded66"} +{"id": "train_09563", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "evaluation_metrics", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a18377aa252fd9b10874009b0dfb8dad207ece41"} +{"id": "train_09564", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdde795cb3e429e038fe7c246f5e3ab6c5d59b8c"} +{"id": "train_09565", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73516b8d55aa000e9fbb3564471a5cb992f508e9"} +{"id": "train_09566", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1d5f0a2fc5a7bd7874ff886af44a45df41543a3"} +{"id": "train_09567", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e57f64ee70d090ba96ab1d1d250663814c95e32f"} +{"id": "train_09568", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca50b7352d18fc5cb6eb519a3b0601249894b5cb"} +{"id": "train_09569", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5747f260949da0d92a0b4ce8e7e392e3c7150008"} +{"id": "train_09570", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "644c54572d966ca15c29a903d92f4ef527d9ef13"} +{"id": "train_09571", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "284ab27e7f3883f53170f1c6d3d4753a46afca58"} +{"id": "train_09572", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b611c34d7509f23f6e68ad8548ad5b9c159da334"} +{"id": "train_09573", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fca39fd0a686fab314141c833e52540ea04b59bf"} +{"id": "train_09574", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e271e8bde7a29751d501a82d6a09953c0e5f529"} +{"id": "train_09575", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "382ae6cd76eb3909cbdf378b3f7290962986704a"} +{"id": "train_09576", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05df00fbfacd78471bb09be97fc93d49556956db"} +{"id": "train_09577", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c75cf6fd47a14fe223c6d895da8fd7607a1639f"} +{"id": "train_09578", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d88a335e02c0d5af42440a578b90e48024f646e8"} +{"id": "train_09579", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "tooling", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "499a9c207db59c5d920a6954b543967e3d6531f3"} +{"id": "train_09580", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "evaluation_metrics", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f63f270b2218f2a8760c10d9f35b04511ee1c016"} +{"id": "train_09581", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "governance", "documentation", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4f826c99e851c51112aa8ae83d81b0c5cd695c0"} +{"id": "train_09582", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "documentation", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "535e1430a648b7e7e42bbc7c1eb057f66e0b8d54"} +{"id": "train_09583", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b36e88c83bc5ad71ccb6d41a83cf5d143d2a0a1"} +{"id": "train_09584", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d4cdb322d76eb3bf8d65cceda4b87c6ebdf9b5a"} +{"id": "train_09585", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d96e0077262d9197793f5c7e262a70a6aa2b390"} +{"id": "train_09586", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "reproducibility", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fc0357265032fa6155d1933ef7eecc8a73858df"} +{"id": "train_09587", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tooling", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d991c6ea4bcd056bc06335b7264f4f87465d1e6"} +{"id": "train_09588", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe9523f8281f043fe9c56172ab706ef3850ed93a"} +{"id": "train_09589", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f3f9253c1294fc84792549d601021af01ef1893"} +{"id": "train_09590", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b524e32a82ec690a976b3fa7769e13cad45fd4d"} +{"id": "train_09591", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "auditability", "reproducibility", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f79f13bd0c3e0d25178890930fad9df6251b33c3"} +{"id": "train_09592", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2537134fcdd164ac33c038ea896758ae1ddec66"} +{"id": "train_09593", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91ced98a839a976238dda1fa807ccbbce637e6c9"} +{"id": "train_09594", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f12e288c8a4e4eb9a57690527f7f1fe993fba301"} +{"id": "train_09595", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cda31a7e6b142bf286fc9840472132f69931f7ac"} +{"id": "train_09596", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3aa2069eb178b66d0c560e72fdbeb274edd7c17"} +{"id": "train_09597", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tooling", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d64c37c649623f50ab92a59fb5c64b6549cc7ddc"} +{"id": "train_09598", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "documentation", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97924175fb9c86b1602e1b375602f6ac7e4931c4"} +{"id": "train_09599", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efffdf72eeb3797829989110e06236b57af95a7b"} +{"id": "train_09600", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63312aabfece2eb9a8d54e010c43bc67e379a88c"} +{"id": "train_09601", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f8cb4bca748ef1ef5e1102f6ae90f41bb48b764"} +{"id": "train_09602", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "governance", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d074c17a0837fd1c7e0dd6d5d90d751d80e72577"} +{"id": "train_09603", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30128394c8295e70a9219f65402f6ca52d54236d"} +{"id": "train_09604", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "tests_are_truth", "governance", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1485357a47f9da9ab843d2ddbd073c82904df452"} +{"id": "train_09605", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "security_gates", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d2a84eb702eb67c43252d2d94d7cb815db255da"} +{"id": "train_09606", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc0c38d6ec6cfb350cb8ee518dad0be203c9c7ae"} +{"id": "train_09607", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "governance", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "976d21d3ee75ce7f17c61255223052609f0c4326"} +{"id": "train_09608", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72742ce671a3521b25afc1a3b396da93775c2525"} +{"id": "train_09609", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tests_are_truth", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb58ac7bd705bd4017989b34aa89e21595c66232"} +{"id": "train_09610", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "governance", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5f46ecd888c55cdae2636b7c10bfae204e01597"} +{"id": "train_09611", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "701187cefaf069dd02fa4b6eca8906897ed971b4"} +{"id": "train_09612", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "tooling", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a73e8a1d0519a25e9cee4d19d1906720fd5852ae"} +{"id": "train_09613", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c42d9ea721b212923d10f73a18599f0a3919a7d4"} +{"id": "train_09614", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "ci_integration", "auditability", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0aefba1b07938e0fcf22acdfd256283306479aae"} +{"id": "train_09615", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "documentation", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddf65f43030d18ebd8de0423e375df5a53b0bd70"} +{"id": "train_09616", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "governance", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6e4706674a3fe3ac124dc7716c0466dcd71e827"} +{"id": "train_09617", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "reproducibility", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e02ca98ab3be24ed7b4d895624d887f3bd5df9cd"} +{"id": "train_09618", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c222c36d559b0e5de74d827659e01d71bea58ec1"} +{"id": "train_09619", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e617dc467b4db70c50538389f118a23fc8b662aa"} +{"id": "train_09620", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5991d16b5b18b4d3544651534b9c92119d0cfd52"} +{"id": "train_09621", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6c8036cc389404d23adfa7dbd8edda1b81f6884"} +{"id": "train_09622", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0313224fe3bdffc5bf6a6ab5ef0e384b7d52df52"} +{"id": "train_09623", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "auditability", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a39666940e7cb5a68304e0e2b58b141229ddf367"} +{"id": "train_09624", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fbb2f13a0aa8b2e5303e0b3a8f2133d7a91178e5"} +{"id": "train_09625", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "auditability", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03104706cf80a4e768bdd597814935dcb5e53654"} +{"id": "train_09626", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "ci_integration", "auditability", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f47d156f4d43251bef55beae06f31b99bb325d8b"} +{"id": "train_09627", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d20bfdc1ab4014700305085cdadab2ae873d556"} +{"id": "train_09628", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc60411947b1685d6fb6139f600f51860a4dc263"} +{"id": "train_09629", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "auditability", "documentation", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd3602fc6f6c9097b94c05c5889c1422c5565c4d"} +{"id": "train_09630", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "reproducibility", "security_gates", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a881723bdc47aa242a457e8178417968829525c"} +{"id": "train_09631", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b69048b4db3c7641b3bbcf4ada432e6e8c843396"} +{"id": "train_09632", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "tooling", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2aa1fc4d151590a18105efe76c63960fbb4071cb"} +{"id": "train_09633", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25c3c312d04b5b3c5f9777dcc8da6522020e26cd"} +{"id": "train_09634", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "auditability", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b79206bc2111b6c8508cb7e6cc3c2f398b7296e1"} +{"id": "train_09635", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "documentation", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bc98b2c772b586e02cc9fa4a14f166c32b37c1b"} +{"id": "train_09636", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef03eb2a1340472394b86a12ec4fae9377771df4"} +{"id": "train_09637", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60deb7e1f227ae109d5fae9b0a8427eccc1c5917"} +{"id": "train_09638", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "security_gates", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b21873863e0e52a0cb3b3d0f7a2b9ec0fe10de90"} +{"id": "train_09639", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8e91cf8eb6a842034cf9055ea6576f490c8422c"} +{"id": "train_09640", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "863363bb0116dc036835424f8a9f51036b513fdb"} +{"id": "train_09641", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f7fe2d3e650bc3c51dfbe087d627e649b5f3b05"} +{"id": "train_09642", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff51666b6458a8b36ba937fd16fd5f1c3c4f535c"} +{"id": "train_09643", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28d170fe0324917376e1b92f7cfd776bf5ae0ce7"} +{"id": "train_09644", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "repo_scale_reasoning", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1a63708361d58bae0588f0979a5758eeb639200"} +{"id": "train_09645", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "documentation", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5457934be61b8b8044e5281ba11e87fae1138b5"} +{"id": "train_09646", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "documentation", "auditability", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2050c203419ffd714678b5f60e382a321f901d5"} +{"id": "train_09647", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60c2be10c9b04549b7a6746210c0e5ac36ac81ad"} +{"id": "train_09648", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "ci_integration", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f9641c4153b30d314fc22ffd8f653ea9e8a01da"} +{"id": "train_09649", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tooling", "ci_integration", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7178dd7bbc96a732ba037defc3eea06e01ad78bd"} +{"id": "train_09650", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d190eb6431d79b9111e05f7352c9a04e03c69802"} +{"id": "train_09651", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b229649999b99b993bc6c09ea11c0de6936027d"} +{"id": "train_09652", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "governance", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18e9fa0a9725715dc9ffeacfa5052b824e8e118e"} +{"id": "train_09653", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f610c852a4140d0237cda85534561b0f7ab8e03"} +{"id": "train_09654", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c8f72078f0d4308309291485a00004de376dd18"} +{"id": "train_09655", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e18b02128574aeb5dbbafef3706f9290be684cd8"} +{"id": "train_09656", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "625088aec6459253d0bc1fcc0ed27490ce2d3f96"} +{"id": "train_09657", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "auditability", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad7c9db38a61093025af857b6dde8ab4ea945e62"} +{"id": "train_09658", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a5f6fec00378c276b31a013e07fb44c7ff84697"} +{"id": "train_09659", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb2fa21fbb15cb91022cd487caa370e8deff4d4b"} +{"id": "train_09660", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tooling", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed2f3a708a93bfde97de6b4835ae8988a8add399"} +{"id": "train_09661", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e846b2423384cbb0e998dde282c0135bcef00c5a"} +{"id": "train_09662", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34b5831028228e32d01c8411779ad9b708d64b3e"} +{"id": "train_09663", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "580a39b7191b2df3db058658581bf7336bc18ef4"} +{"id": "train_09664", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "governance", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e75face8f95eeebd2a17815399c7cf8177b1bf7e"} +{"id": "train_09665", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "ci_integration", "governance", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b50c63cb63617c605c7cda73e611fe1cbe08ec79"} +{"id": "train_09666", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "governance", "reproducibility", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f150a9f5df005d161029d115b42060096999cbe"} +{"id": "train_09667", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af2b6e38d7ad66803bbafa7d47aed277a272a78a"} +{"id": "train_09668", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7d60c1562a3ea10ead990825479271f6a71412b"} +{"id": "train_09669", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "evaluation_metrics", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "32723a0ca2376f8e66d890b2d1cb75430a7558bf"} +{"id": "train_09670", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18a1f6e4da83ceef4529784ee0ab29b35b58d766"} +{"id": "train_09671", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "tooling", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "205e97396bcd31f582aa051a072f47874380fb7f"} +{"id": "train_09672", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "auditability", "governance", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e16b164e9c39f9f4b8279e86979abe0bd3ced9e5"} +{"id": "train_09673", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5877ba78a29d27adbb86ea932611c56ccd5ee781"} +{"id": "train_09674", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "431cef9db010f2141ebfe404d08c0f8ef8a80d3a"} +{"id": "train_09675", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "documentation", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "648933285099b49f2bd831ce0c428188549d2306"} +{"id": "train_09676", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "governance", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a54b8aea89c0f329eaf952431aaf509db525b320"} +{"id": "train_09677", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79e19485a73e135e1e9fb6c265fa115e92b459eb"} +{"id": "train_09678", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae9d24a9fe4e35b17afef23d141055171863ed2e"} +{"id": "train_09679", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3516c3dde1c121aa350e854518184f011105bb74"} +{"id": "train_09680", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d1a15c307d51a89ff64460c5fea8793f5435225"} +{"id": "train_09681", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2edf0d259227290dac65b1c1145a62173bc36fea"} +{"id": "train_09682", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "security_gates", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8a5526c1ed8705f1deeacaf17f0275390adc804"} +{"id": "train_09683", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "ci_integration", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6af45377eadaac268716343f3db6393b8e31a338"} +{"id": "train_09684", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f9560f9953f1e15ce4ae13acc0ce785a4066a8a"} +{"id": "train_09685", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "tooling", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b291485fd5ed8543ee8f95d083d8a7621fd685bb"} +{"id": "train_09686", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14c3aac3a49f680411608584a48811f0bb950296"} +{"id": "train_09687", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "328c4eb47787808e6e2b23c696da93c1a767779b"} +{"id": "train_09688", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00b8c37535032a7546c3f78118a34a80a926e8e7"} +{"id": "train_09689", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "666c3f35a5766c1ee96ed0ca98e5f00f87a505ac"} +{"id": "train_09690", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "465b93556b8086c19cb3134c507ebda74aee5b9c"} +{"id": "train_09691", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e0ac8445a79ed94c507e5626c557a875f0ac124"} +{"id": "train_09692", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab2901dbe12a8494cc0e51ae4650f2759acef548"} +{"id": "train_09693", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3310479a7f3dc342ed8c36e1b2d4b82644bb188"} +{"id": "train_09694", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1455f6c75fd10748ecd4433dc84e9aaf92550df"} +{"id": "train_09695", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b25df752428455c15b5507f0b913f8e7b9290c6b"} +{"id": "train_09696", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b97134354e80d1227e7161fff7f09b429a8c71c"} +{"id": "train_09697", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d42831b6e9bfd09597a62b1cbca030bd6cdb2614"} +{"id": "train_09698", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "589cea79f06b30bf029e112cb4d7d27d77c646c4"} +{"id": "train_09699", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2531974312a81dbe7f6475d873732f4ff049349"} +{"id": "train_09700", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "auditability", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9f75d5829264fe3f69eedd2f92dedb182531074"} +{"id": "train_09701", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "governance", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be070ead69d7bab644913d5d4b55bd2d95278a73"} +{"id": "train_09702", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c05ef12fe633dec866a5fcdeac40f8d92cf5c0e"} +{"id": "train_09703", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e1b00da5daa5143cf9ed967fe3abee5bd4aa520"} +{"id": "train_09704", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4c5be0fd1a32a1a55d8c273520b9eaeefdc4581"} +{"id": "train_09705", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0f7a3a6c64ce029f1aedb0b93285a6e3ea7a038"} +{"id": "train_09706", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3e0a84cecac7a1e29211df62adf6061b7beac6f"} +{"id": "train_09707", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "security_gates", "ci_integration", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "630bbcf205efc6c2d03defe8f8552c4445a973d0"} +{"id": "train_09708", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93d2456f0b762c01511b66aedc6a20a0051b2491"} +{"id": "train_09709", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "governance", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4244ca7fb3b22f542c2539d14ab0b8033101a493"} +{"id": "train_09710", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "auditability", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0748da530f25c1bd70021467f74f19070e520e17"} +{"id": "train_09711", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdef52536e5de16e223a9f1ea54229263934e416"} +{"id": "train_09712", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "tooling", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea3c83a361065eda66986f472aa18b17842cddc3"} +{"id": "train_09713", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78249f821dcffa92a3d1ff9138a5398a5367f08c"} +{"id": "train_09714", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b37e311984abe34954b5188f8bb6e54fd8b7a054"} +{"id": "train_09715", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efdbe375940aad0b7f7d86a7e45b5d8a80a22270"} +{"id": "train_09716", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12886fc7bcb6afc74fdc47092ee07e4c853dd4ca"} +{"id": "train_09717", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f93cd7f2c3ad71cb5c7c2a79d90352f21ada4eba"} +{"id": "train_09718", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9fa69f5ce0bdea529abc7c95bdf416e124acdcf"} +{"id": "train_09719", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffb7a67a41e84e1d6e3f49b66fe61418b56f3aaf"} +{"id": "train_09720", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "092a10e3e539972ba80843a80c36ccdc4021669e"} +{"id": "train_09721", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b54ba475728a6fcbc158dfd726bde9b1774e347"} +{"id": "train_09722", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e9b5877b611ca1199ca7956353c7ffd5b3ffb4d"} +{"id": "train_09723", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "governance", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84d899f048aeab510664989f128bb871802ce1ae"} +{"id": "train_09724", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "security_gates", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56ed35546b96e53627d595dd62d20768470187ba"} +{"id": "train_09725", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5d39326fabbee6cfd48b5155760135c126fb1b8"} +{"id": "train_09726", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "ci_integration", "governance", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fa0494d304ea6aebdecdd188b2bae3ad6d31b94"} +{"id": "train_09727", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b0787e6ba400121b82e4e04c2a7e33de97ce81e"} +{"id": "train_09728", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "security_gates", "documentation", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef985048ee27a820a8a2856c0e9db98b9ded104"} +{"id": "train_09729", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22184ebe96d6a0b262273756135b9cfcb3e401e0"} +{"id": "train_09730", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "ci_integration", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "051d64dc518573e33af39fbf3ac2853b2daaf05f"} +{"id": "train_09731", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd8d085ec027e19e19e2dac7a8f19d07798bef52"} +{"id": "train_09732", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f73b4c022a3b1b99def755fd74a2562a466e7f9"} +{"id": "train_09733", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbb0098d076b083686201037ed107f01b6e2475b"} +{"id": "train_09734", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "governance", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad5fb4c5c22bf81347684b7d1f4ceb3d7b5c0432"} +{"id": "train_09735", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "security_gates", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "987f64161d352c727afd814d8878c7eb0592fe1e"} +{"id": "train_09736", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "security_gates", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dd44274fa2df2f9a5b3559b088fbf7d2faf5552"} +{"id": "train_09737", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53510ee5ab4ab8d1971cf0e0839b7aedf9a6ff49"} +{"id": "train_09738", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70aec5b7194ffb925ced89bc0a278a4169b9c684"} +{"id": "train_09739", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "governance", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b307c2c113faf755986318b1803efcbe52f5498e"} +{"id": "train_09740", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1430a62d71fc4fb7a744e16ede6f4318e0e21aab"} +{"id": "train_09741", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ad213088ebfa4db02c4b9df9d1c1397abd0a7b7"} +{"id": "train_09742", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "auditability", "tooling", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef9f54720b38139fb593f707e8d609f875e39234"} +{"id": "train_09743", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "497ed367a3dab335b8ccfbe5dedaf111d78902d9"} +{"id": "train_09744", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tooling", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d1920fb46a2f44d904f1906a3eada5ccd9ee07c"} +{"id": "train_09745", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90f714ddca20ddf275703e3073d223437e1e62f0"} +{"id": "train_09746", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "534c4f79562f554b90ad3e3e094b733664fa003e"} +{"id": "train_09747", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e64cec33eb13e61bce3e09c0b3bee1cf72fe559"} +{"id": "train_09748", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b59792a66f35d80d6f7e07bd539b5ae7f1ee9d67"} +{"id": "train_09749", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "governance", "tooling", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56aa9f6fc91786642ddcb00d941578180d61fcf1"} +{"id": "train_09750", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tests_are_truth", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a648726772ff621bcb08175eb331b1fcc8b2ae0"} +{"id": "train_09751", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "governance", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2ac8235c474f1232b8ec6e1c29f5a7c871148f0"} +{"id": "train_09752", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7389093a43570315b1b73f63c5dcadb1803579c3"} +{"id": "train_09753", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c41d2f44a0d5f8eec8ce0fe1864831bfd4be248c"} +{"id": "train_09754", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56aa9f6fc91786642ddcb00d941578180d61fcf1"} +{"id": "train_09755", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1bc4eb6c3862f829116c7c63e568fad0f7a7f113"} +{"id": "train_09756", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "tooling", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "321d4cfcd9e0918650f77b9be4790a2cc744c493"} +{"id": "train_09757", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "evaluation_metrics", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "febb6c86f1d15ccd05990cacf9438e06006fdd44"} +{"id": "train_09758", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "810e985fb1b62da311c2962ef421fdf45a0ba92b"} +{"id": "train_09759", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8bb467fe16f0393caf8096ce3c4346b105140ab"} +{"id": "train_09760", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "169739bd905384d85b85f1ff825603d808a164db"} +{"id": "train_09761", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9c3152e5e1bb71fe3f5b42f415fd911ace15c1e"} +{"id": "train_09762", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "documentation", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f50accf8b7edc0fd9e507b605ee048bcb3554750"} +{"id": "train_09763", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "governance", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "780ba11b521df7e73a97ead9393aeee1444e5e1c"} +{"id": "train_09764", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "auditability", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5798d5c5c3ae0b37846ac6df8bf5b987134626a0"} +{"id": "train_09765", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a36d5375b64a9686abd0dc5dd8868e30cc029005"} +{"id": "train_09766", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53dbb932fcff126220cd972bc3d4e5026bf25e6b"} +{"id": "train_09767", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45a744e6246cd97062356caca90773660723f104"} +{"id": "train_09768", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "documentation", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "320b5b0455526ae493e9419f3b637e8b290385db"} +{"id": "train_09769", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "auditability", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6633f6eb00415a3bc6e98079466feb79e90c715"} +{"id": "train_09770", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6af6ee04ad271abec9d68cd7a6a8b7bf93fe417"} +{"id": "train_09771", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "reproducibility", "documentation", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff9260c4463042daba1689aef43c9133cc740918"} +{"id": "train_09772", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "documentation", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8c28525570c3b36ec029370ca1e5d2120fc8219"} +{"id": "train_09773", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a78d65a3528b2e1bc19dfa868775c7bd206e8e4"} +{"id": "train_09774", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "documentation", "governance", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "178d58c7ff7a172c0817cb44d95a3e3e33e526b3"} +{"id": "train_09775", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c5f1bdf8fb4f0879dac7cc24c8213e37267101a"} +{"id": "train_09776", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "evaluation_metrics", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b92795b778d1fa96c41f0c317e589334b98bb49"} +{"id": "train_09777", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "reproducibility", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42bf0b6efa0a1261e47fc2d59f52245069189811"} +{"id": "train_09778", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f610c852a4140d0237cda85534561b0f7ab8e03"} +{"id": "train_09779", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "repo_scale_reasoning", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f87b5c9836d99fcb8cc948b33966de0c18d56c13"} +{"id": "train_09780", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "992f9d13724fc1c7761538c825d81e3dc3afeb38"} +{"id": "train_09781", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "872effa4de644390e02659809f67d30004f17ce2"} +{"id": "train_09782", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd298fd15ba61c7dc8b967a256c9f4c31305c638"} +{"id": "train_09783", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0eca96a5362f57683f2078090c75292e7eee95f"} +{"id": "train_09784", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f058e1badcd4bbc176b9392029c6a4104ab9cc8f"} +{"id": "train_09785", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ad0138928ecfbf4fd0a75786d3862d85c7af114"} +{"id": "train_09786", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92a5bc9f74e9370469a3ecc24ae3a459736e5791"} +{"id": "train_09787", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "cost_latency_tradeoffs", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "862150dcf38eb6803f764de0cae2f7f1ba1dfc4f"} +{"id": "train_09788", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4573ba38159631f00c6a03b89c32cf4fa08d8e0d"} +{"id": "train_09789", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "evaluation_metrics", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a8b17b1cb74705fb2940a0c8fb16a70a73d2581"} +{"id": "train_09790", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5b262ea8a1978dc1c179d00dd9a05ed292c9c2c"} +{"id": "train_09791", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5f986eb4dcd9240b574df50af57f7c829e9f5fd"} +{"id": "train_09792", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b98bab65dfec8814fe25dce736b760c463c2dd8f"} +{"id": "train_09793", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97b3b25ae9e981c12f21e9c37cf2af9a15c52788"} +{"id": "train_09794", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b565c64d0c4a70d116134a166ed3ca30a9a709ee"} +{"id": "train_09795", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5659aa78b9e66138af0df532f3526fcd74ff3e7b"} +{"id": "train_09796", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6592e54f54b2f43c740e4608ecadf754ccfa178"} +{"id": "train_09797", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "reproducibility", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8408c38f3276a5e81b338690d90f1b587341f9ef"} +{"id": "train_09798", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tooling", "security_gates", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c559a2d943904d0308d36cbe109ab198c9eb2056"} +{"id": "train_09799", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "reproducibility", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6075d88683d53f041756c48b9826a62209bd4232"} +{"id": "train_09800", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nScaffold:\n```python\ndef agent_loop(plan, edit, test, issue, max_iters=4):\n history = []\n p = plan(issue)\n for _ in range(max_iters):\n patch = edit(issue, p)\n ok, report = test(patch)\n history.append({\"plan\": p, \"ok\": ok})\n if ok:\n return patch, history\n p = p + \" | refine\"\n return patch, history\n\n```\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "security_gates", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d722b93bbfe7c2d8cb5914a96508b5cbd11c1a47"} +{"id": "train_09801", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ead3831d6643d2c1fcaa1f84e57da28f7c7e50d8"} +{"id": "train_09802", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d226c60180865770d7f54be20d3cca8a053b10e"} +{"id": "train_09803", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "documentation", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e174230e1bfdfc8e4f0fa465ea21e8b77460d5fe"} +{"id": "train_09804", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tooling", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1c1d1a564401eee99768665e8b3a55ae3250d56"} +{"id": "train_09805", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca69145d719ffceeccac3115fdde19eb2bae28d5"} +{"id": "train_09806", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "governance", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a6a931fffd7071107bf73e33b92e4e73d9df778"} +{"id": "train_09807", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tests_are_truth", "ci_integration", "governance", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9478758d2f704cdd46224563f4d1b821aac87de0"} +{"id": "train_09808", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e7b5f0ba223b699c45ceca831e311e0803a5de1"} +{"id": "train_09809", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c09017d346d9b7c5eeac9844919ad1beb476453"} +{"id": "train_09810", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "governance", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "725f49e764b57aa32dd513fda0b59225d5c20533"} +{"id": "train_09811", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63e379baef6e624bf6454fe1f425e378a7ab7be2"} +{"id": "train_09812", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "auditability", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59f3d322de8f7b9deb1a468c750ce1494729f510"} +{"id": "train_09813", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "reproducibility", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "746bba967e00116397eb195318d009b69dc979e3"} +{"id": "train_09814", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5aa77c02c033bdad1ecf3d1a788c2d693e79bf8c"} +{"id": "train_09815", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "596f0698d908f052a917ef77a22216f9078ecb15"} +{"id": "train_09816", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "governance", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "509acf19c89aa304ca560ba8f44ff4663848223d"} +{"id": "train_09817", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "reproducibility", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b702ac911ad605c1078f4d7d90f54be9e4a6a0a"} +{"id": "train_09818", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc2485ae7eaf55b2bcef59c724b90ba059085711"} +{"id": "train_09819", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0ba3d34319b5a674e964f0fed48a697e98d200d"} +{"id": "train_09820", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b68dd0f37c27fcc29356f00ab9206b84ebce4e28"} +{"id": "train_09821", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ede2f21e135c847c3718a1120440138d7b9c9fd"} +{"id": "train_09822", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Python\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb19004c47e5027b7bd2b6e3573a6df5108f664f"} +{"id": "train_09823", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "governance", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "964d7c9f360dde82d78ca2a2868ca3847cbd65cb"} +{"id": "train_09824", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d44a6604c82c258b09fddf2940a4ad5209a2f680"} +{"id": "train_09825", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c846bc5b10ffaa843ffd7a6feeda79c4a15544f3"} +{"id": "train_09826", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fad51c712826ecc3aaba0510caf5374bd882ce21"} +{"id": "train_09827", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "repo_scale_reasoning", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f91a09497b621d12dbf2e24ba35ba39305a5e8bc"} +{"id": "train_09828", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ae0fc400f2e8eda15c2b3a0475f58b0caa95b81"} +{"id": "train_09829", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tests_are_truth", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8a8f907e55b5b935fb47a21bc8b65a6daed9f9d"} +{"id": "train_09830", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd46084b51062e2b72925369c42111dd742e972b"} +{"id": "train_09831", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d016895a29bddd974c985bb100abed2ba3e75e0"} +{"id": "train_09832", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7f9adee30364074f2631100f36d757930959c37"} +{"id": "train_09833", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09a6e4b90797ace3bff878d7c180aeb62f750369"} +{"id": "train_09834", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e6317c0dd0edd95e23cc51fe226632da848f332"} +{"id": "train_09835", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2240c130e692cc7c383e0e7f98dcbcff4da4744"} +{"id": "train_09836", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b01d2f4cc79e2133c3007bc5fc8277fcea48750"} +{"id": "train_09837", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "tooling", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c69feca5743540df01f0c8321fbf6f69960ba61"} +{"id": "train_09838", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64e8d22feb3a1901500cd00f5ee7acdb10c1b4ae"} +{"id": "train_09839", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb411324b0d265e5780e540c79c8ff41f0cc6ec9"} +{"id": "train_09840", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "evaluation_metrics", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ac8f8fed6703f01dc6b8c24a8bcc16ed39668c5"} +{"id": "train_09841", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c87afe199cc6c76ea4af09af2ced1e9dc595af0"} +{"id": "train_09842", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "ci_integration", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db39747f167e8ddf78e927ac28d8b00c494ffde2"} +{"id": "train_09843", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47a3095756bdeb6a011802700ca47d4258d78ff0"} +{"id": "train_09844", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4615d90119f5fc94507145e43cdbf3f3a1cde40"} +{"id": "train_09845", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af47f786a2f62667f44778f51b47dcba61559ae0"} +{"id": "train_09846", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6a4420c3abaf22b63d2e5661d64dbf73c120728"} +{"id": "train_09847", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0fb01a19fb7295182706fcce6e88f7c4c56d2aa"} +{"id": "train_09848", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "ci_integration", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b37bc8ae873dad27500be6b23317f466e8c0263"} +{"id": "train_09849", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "governance", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2f1809636967336bb09d729e6b684dc9265fdbd"} +{"id": "train_09850", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: TypeScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "093cfd1e8e2b409a3333eab1b95658a13c0baf36"} +{"id": "train_09851", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "cost_latency_tradeoffs", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1055eba9ad97437efd936a4212791b98883d0a4"} +{"id": "train_09852", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7583497c97360104f2198f228833cc36e221dac2"} +{"id": "train_09853", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dafbd3f36642b2576cdfcaa621bc55a02259298"} +{"id": "train_09854", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Python\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tests_are_truth", "governance", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c047e725551bb29efc77cd10c918b7d1b48f6ca"} +{"id": "train_09855", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ddd235638310c307df6f638d71b922793f0449c"} +{"id": "train_09856", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "112f6f914d6309814c6e25adc2dfa80608a9f908"} +{"id": "train_09857", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "auditability", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "147b122748e928d7ec24c864c5225673778d6b58"} +{"id": "train_09858", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "429020ec1bc9a558ea75c6f4fa49ae5dee2fff3a"} +{"id": "train_09859", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28f2c178363fb7b9a5d8e991a9c6891915899556"} +{"id": "train_09860", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: intermediate\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc437405fcc457f1de769a4b9b5a42004727b5d1"} +{"id": "train_09861", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6e0327fe735d1c66fb9e74229e9e41db155ab41"} +{"id": "train_09862", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tooling", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0b2634b82d127ec9ec3a05dd90eb72a4e839a4a"} +{"id": "train_09863", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a1b4048e860b27d5279cc7da6a6298c34e70de9"} +{"id": "train_09864", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01e42daea840880e6c2315b2a293ecb4a8fe9e95"} +{"id": "train_09865", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Java\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "tests_are_truth", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ec057e84a3c4632974ceaaa9a296188b452450d"} +{"id": "train_09866", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bb00640aa59420b8bb9d34bfee8aaa5cf172d01"} +{"id": "train_09867", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c52ba51cf9742672c691660acb9a14afb3d8267e"} +{"id": "train_09868", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "auditability", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c35ac634f30646b886b51897ebabc6752a5629d"} +{"id": "train_09869", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "cost_latency_tradeoffs", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c43f008023ac5502f461ab1c70821ecbf5eca79a"} +{"id": "train_09870", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: SQL\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "documentation", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47d96750360083eb693a6908569ce2c4206bc73b"} +{"id": "train_09871", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "documentation", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3be35998aa1177c74ed0ca3ba3834c2c18a45e4"} +{"id": "train_09872", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf601cec3704a054a53605741e1573e9bfd5bd27"} +{"id": "train_09873", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "488a0b3281011bf8533f7ac504b32219d2da5bd4"} +{"id": "train_09874", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae8411a795d69207c0953bf58cbad41f18c5156c"} +{"id": "train_09875", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a48d85509cc98e6c0247c62339e59e73c05f87bd"} +{"id": "train_09876", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "tooling", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a7d117f046270288f416902d4cb333f0f7fe9ea"} +{"id": "train_09877", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tooling", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1c926ea1a1fdc13feb9fbbe3d42d4289578efc1"} +{"id": "train_09878", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "ci_integration", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2faef4557781a5c3236709ee6f8a94352f04a0e3"} +{"id": "train_09879", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "reproducibility", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0708930087b4e77ae2439ad3cc469b4a227b8e09"} +{"id": "train_09880", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34c5122b6ea3ebf4d89d152ec30bb3b0ba1b3602"} +{"id": "train_09881", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56fde241b0fee8f1b74264301ec7608f9ab2f03b"} +{"id": "train_09882", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc789a1afc67b390f4eb323429afeb01a6b33862"} +{"id": "train_09883", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "ci_integration", "documentation", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9dfeac8ec310bf9725fc9c96dd002e8510e1d26b"} +{"id": "train_09884", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f334bd1924227896206c15d0abe6ee240ed0804"} +{"id": "train_09885", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: expert\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfaf866577e7e73d0962ae0d0e03ff70bd5c9ac7"} +{"id": "train_09886", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31b70315c7753574af5243747a0196d0182d4470"} +{"id": "train_09887", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "documentation", "tooling", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dede9eba7d2283e8eb936d62b3fb96733cf8a97b"} +{"id": "train_09888", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["governance", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5f9f4c76a9d8af65b4075e58b78a9c979b4aa61"} +{"id": "train_09889", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "intermediate", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f311ce2d4b70cd3b4c664028609ddba57cb7698"} +{"id": "train_09890", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Rust\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "reproducibility", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72bfb6f57bf2df0701e45b84e32a4f61f227348f"} +{"id": "train_09891", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tests_are_truth", "auditability", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b22725a92aa036ed829cd7e10ff84ae09d7dd8a"} +{"id": "train_09892", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f171cb3a6b44e61986a3bc1742d1331e6ed1832"} +{"id": "train_09893", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed1467431872b97b11f67edec706bfd2b008612d"} +{"id": "train_09894", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "ci_integration", "governance", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c4177cf3a63c6cf516d66244f83011c977c0811"} +{"id": "train_09895", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae05f8f92feebb497c9e07dd36d6085502be6d64"} +{"id": "train_09896", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "explain", "difficulty": "intermediate", "instruction": "Task: explain\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "tests_are_truth", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ece35f177959a0537267befb4d28ea852a4804a"} +{"id": "train_09897", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07d2072bd64a8b269317df61b9b0c7a8174c1801"} +{"id": "train_09898", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: SWE-bench style real-repo evaluation\nDifficulty: intermediate\nTarget language: Java\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6ad07354aae5e73ab62bd14438bc6ca7985f81c"} +{"id": "train_09899", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a455db932c82e5532d80a79fac02233be94b7c6d"} +{"id": "train_09900", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "auditability", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48d33959b0e7d343329e3834f63fe1fbbd9211e4"} +{"id": "train_09901", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8045e5dbc54084a77c6b22a83f4a898153183931"} +{"id": "train_09902", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e4f70216048788a37f984ffca286baf0786082e"} +{"id": "train_09903", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "intermediate", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "tooling", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "482e1c978a898ecad89b70adfe5a74b1028979fd"} +{"id": "train_09904", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "governance", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d279fea0fdb8c5d426381a15007d6b79228b049"} +{"id": "train_09905", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "governance", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa9fe2d9ce0ccef9e21634a2ec68658e817e8bcc"} +{"id": "train_09906", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "documentation", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e18432b2ac01149f11f5cb2b0fb3bb8616675727"} +{"id": "train_09907", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44ae5cb5e3c4669f0a095a46913905e9a383571e"} +{"id": "train_09908", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "reproducibility", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c22da178fff110e96e0ce1d1c6fe46b944495fb"} +{"id": "train_09909", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "tests_are_truth", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc47c06cf24af6d938547b59f8b2645aee812b1d"} +{"id": "train_09910", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "security_gates", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8469aeaf1cdfd6b81ec0e5a447f7fb72b789702"} +{"id": "train_09911", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "680dcc25f2cb7ed6307bf4b8400008932a12fd2a"} +{"id": "train_09912", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "governance", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3554b25225410faceffe0a9cd53db5cbd931fa8"} +{"id": "train_09913", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf101338647712e0447089178065c97188789103"} +{"id": "train_09914", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "980d6c70c88f84542dfa13aaa7ea584cc331675a"} +{"id": "train_09915", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04f1ce4e92d6164a6bc03bce8068f84f9474ffa0"} +{"id": "train_09916", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3774de09a4482467a43ee00f34f990ec261f3536"} +{"id": "train_09917", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de9c6d193bdd6de106c62234b37bda3c565fa005"} +{"id": "train_09918", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd2e0671324624cd71d61ff66e905274d74c56c5"} +{"id": "train_09919", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "reproducibility", "security_gates", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e75d1b57871cc4786405ade984e038707d0aaef2"} +{"id": "train_09920", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: expert\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "reproducibility", "documentation", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25204e6559c21bb6cb760fa2cf93c5817f023f7a"} +{"id": "train_09921", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ccc119b3cde2e71cc00feaec48cb91d8d870c2"} +{"id": "train_09922", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "auditability", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b43b2833f2668ed5bdb9ed9c9c71dd15fae53efa"} +{"id": "train_09923", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "agent_loop", "difficulty": "expert", "instruction": "Task: agent_loop\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: C#\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d247ae5bd6c495f35a36b77d9badf7671d43aa4"} +{"id": "train_09924", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee9c2efa0cad28cc1c8d8036bcfc956254ce75ce"} +{"id": "train_09925", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Bash\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "auditability", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dba8452eafbb660ddf0bf35eb15f5bc593401054"} +{"id": "train_09926", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "compare", "difficulty": "expert", "instruction": "Task: compare\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "documentation", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "136443340a458ec290aeb715b0b29aba2796dd64"} +{"id": "train_09927", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: intermediate\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "governance", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab78223034f1b449bab5f879d6260468d6b5c418"} +{"id": "train_09928", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "tooling", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69302c240fd0ccca6d084b4a907877c45c8a4a1d"} +{"id": "train_09929", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01a9a85ba5814133fd1bab7ced2108d6baa1844c"} +{"id": "train_09930", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Python\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f0f57fef0a056d8d639b4122e2366f183aa1ba5"} +{"id": "train_09931", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "documentation", "auditability", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a8ddbf54da067197e9215a666babd6cf4c54b60"} +{"id": "train_09932", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc72c192bae800a8215212cc82b57ea7c09ea2a2"} +{"id": "train_09933", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Java\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "549d004828fb7c1898b065f8b9079820784fe6d9"} +{"id": "train_09934", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "compare", "difficulty": "advanced", "instruction": "Task: compare\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f309d769211b31ce082fbe9b7bc12e3c7f34da7b"} +{"id": "train_09935", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d15b6df9968abdc112b55a39361628ecd795d507"} +{"id": "train_09936", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: C#\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "governance", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b54cc74b9a408a898adb3b78cd5e3987112e8880"} +{"id": "train_09937", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Latency, cost, and reliability optimization\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9eba037f5528472a0bf10537dd2a71041cb1c204"} +{"id": "train_09938", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "152a4664abc282d29c295fb47a8fb0137f49a85e"} +{"id": "train_09939", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40f8d017ae7a95b019184d268dcdcf2d426441c6"} +{"id": "train_09940", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e16d9dee6dfef15cfd0fdcb7be00ecf1b91c72f"} +{"id": "train_09941", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "security_gates", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82eab93b9b93242c0b57bff4e6e528cb0d6c5efb"} +{"id": "train_09942", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "design", "difficulty": "expert", "instruction": "Task: design\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: Go\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dec4185e24f66bdbaab61e6ec3790a6809c8698f"} +{"id": "train_09943", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29ea5543236ac4535342cd99619742063b7beedc"} +{"id": "train_09944", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3065d4909ab721d75371759ab2bdd5fac7cf127a"} +{"id": "train_09945", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: expert\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d772a6290b841a6b4b4882b6058b5c5fb048a425"} +{"id": "train_09946", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: intermediate\nTarget language: SQL\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee72848198820c0a7617ad9e9ebe402bdc49af7b"} +{"id": "train_09947", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "agent_loop", "difficulty": "intermediate", "instruction": "Task: agent_loop\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: intermediate\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c821f7c45f697cba22a5d16bea5e327f2df1d404"} +{"id": "train_09948", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: Go\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "auditability", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4568a6a944678b12a5cf11e1d270ab29b1edeaaa"} +{"id": "train_09949", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: Python\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef96c06b568f12dc72a6ca5eec4e81085e595f80"} +{"id": "train_09950", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Governance, provenance, and licensing for code data\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["auditability", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09b2c2f8aa60b0021373f2fbd0482e5d999117c2"} +{"id": "train_09951", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "auditability", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f83a36c861dc3c1afaff45dec783d7804d29ede5"} +{"id": "train_09952", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "governance", "documentation", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "216912256d04bda87a6f5c056010a9d8273a51bd"} +{"id": "train_09953", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tooling", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6f90a393e4caa8b28cef784bf28b3a2c71e54c6"} +{"id": "train_09954", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: TypeScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "governance", "documentation", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "227d194bc8be4e6935ec48e074ce92a90ad212bd"} +{"id": "train_09955", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "tooling", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcddeae68574cc57640c4e3c3f772cda7103a735"} +{"id": "train_09956", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: SWE-bench style real-repo evaluation\nDifficulty: expert\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "security_gates", "governance", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c0cb02d2979ef213dd4ad360239e4eb79801b74"} +{"id": "train_09957", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "compare", "difficulty": "intermediate", "instruction": "Task: compare\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: intermediate\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nCompare: capability, cost, latency, reliability\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8daeb0b0dc212ddff55e86db3360c77869af723"} +{"id": "train_09958", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "design", "difficulty": "intermediate", "instruction": "Task: design\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "security_gates", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85105b0ed13ce03c811aa18c38a41708c3c0450a"} +{"id": "train_09959", "created": "2026-01-01", "topic": "Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Agentic coding systems (plan\u2192edit\u2192test\u2192reflect)\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82a76af5afc4b041a30f6677a22a609b7622ac9a"} +{"id": "train_09960", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57ea524050525cb2488a07af3e10a1eaef9418ce"} +{"id": "train_09961", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b12bf8015d31e13ae0a8b7d343cac6c6ff3ba51"} +{"id": "train_09962", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Self-improving agents and feedback loops\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "105a42a98ed2829cc300d9147d70d9ca653b8765"} +{"id": "train_09963", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "data_pipeline", "difficulty": "expert", "instruction": "Task: data_pipeline\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41508e594d4a29037cb9af4b56dae65712e7bb5c"} +{"id": "train_09964", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "security_gates", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3e3a87ab9d3593fedb12c197be5de81ee8b2fbe"} +{"id": "train_09965", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: advanced\nTarget language: Rust\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f47d29af4e3c920b63fc7455dcb7eb632a0918a0"} +{"id": "train_09966", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: intermediate\nTarget language: TypeScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "documentation", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d895ee385a3c6c1ec673a75632c00891b99ef79"} +{"id": "train_09967", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "ci_integration", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9dcf720f523e9a761e0c813d5a9ab087e78158ae"} +{"id": "train_09968", "created": "2026-01-01", "topic": "Mixture-of-Experts (MoE) for code", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Mixture-of-Experts (MoE) for code\nDifficulty: advanced\nTarget language: Java\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9e23de698b1d2a4a3fc33ee3ffb84b634a7c83b"} +{"id": "train_09969", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "failure_analysis", "difficulty": "advanced", "instruction": "Task: failure_analysis\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e20a757be83fcb99e60d83c845c084a5b3401c7"} +{"id": "train_09970", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "agent_loop", "difficulty": "advanced", "instruction": "Task: agent_loop\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nLoop: Plan \u2192 Edit \u2192 Test \u2192 Reflect \u2192 Human gate\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87cbbb13d885cd82396858f95f1328ebb0f2c20c"} +{"id": "train_09971", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "reproducibility", "auditability", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d0c3757515914d58a7f5c63112f047c9677ec27"} +{"id": "train_09972", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: advanced\nTarget language: Java\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Java", "developer_needs": ["documentation", "reproducibility", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7550deb1cc5f93e94e368506ec81e522ab055690"} +{"id": "train_09973", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: expert\nTarget language: Rust\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "858fcb83a80143cbf70c1e05cc675de61d72d7d4"} +{"id": "train_09974", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3709b4cffec0c76f1cac074cf25d2f2683a023f2"} +{"id": "train_09975", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: Rust\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "governance", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05fb6bbb9c2b5551344bec8f5d454b91268e852d"} +{"id": "train_09976", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "explain", "difficulty": "advanced", "instruction": "Task: explain\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: advanced\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae2186d227cf86cbbd88f90129b5dd6eff24377c"} +{"id": "train_09977", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "data_pipeline", "difficulty": "intermediate", "instruction": "Task: data_pipeline\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: intermediate\nTarget language: Go\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a74f9af72f6831c2eb9db70b7a766243d8ba9fe7"} +{"id": "train_09978", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: expert\nTarget language: Bash\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55603140465e5cfc35176a0044ca828a25619e18"} +{"id": "train_09979", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: SQL\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc7b499f5a60b7ca224d4051c1e3ab96dcf39045"} +{"id": "train_09980", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "patch_diff", "difficulty": "advanced", "instruction": "Task: patch_diff\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed57623a5e5dd930391385757866726079a88402"} +{"id": "train_09981", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "advanced", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12429a7ea054f311c6dd3712225dce7853d3193e"} +{"id": "train_09982", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "expert", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: Rust\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fbbedef2e55f3fa5b8d85b7b8eb97af0289c53f"} +{"id": "train_09983", "created": "2026-01-01", "topic": "Extended context and repo-scale understanding", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Extended context and repo-scale understanding\nDifficulty: expert\nTarget language: JavaScript\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "ci_integration", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd0b3e5c21108e4dbf09c30830b468f85cd6ea47"} +{"id": "train_09984", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: JavaScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3768d71cd78f2e0cd11b4c8f85d513cf75b45017"} +{"id": "train_09985", "created": "2026-01-01", "topic": "SWE-bench style real-repo evaluation", "task_type": "data_pipeline", "difficulty": "advanced", "instruction": "Task: data_pipeline\nTopic: SWE-bench style real-repo evaluation\nDifficulty: advanced\nTarget language: Bash\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPipeline:\nIngest \u2192 Normalize \u2192 Filter \u2192 Dedupe \u2192 Quality \u2192 Mix \u2192 Audit\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "documentation", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b674a37da16fb7f860c6a083094785e985a302a9"} +{"id": "train_09986", "created": "2026-01-01", "topic": "Code-specialized model families and sizing tradeoffs", "task_type": "review", "difficulty": "advanced", "instruction": "Task: review\nTopic: Code-specialized model families and sizing tradeoffs\nDifficulty: advanced\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "security_gates", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0cee180d601c2951f056633f30a4f527efb9f84"} +{"id": "train_09987", "created": "2026-01-01", "topic": "Latency, cost, and reliability optimization", "task_type": "review", "difficulty": "intermediate", "instruction": "Task: review\nTopic: Latency, cost, and reliability optimization\nDifficulty: intermediate\nTarget language: Python\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddcd471e6318b2fcdd152fc7c0b5d8ee84e3ba01"} +{"id": "train_09988", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59e572769cb84d17b227ef8a26d9d99d912d16d9"} +{"id": "train_09989", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "eval", "difficulty": "intermediate", "instruction": "Task: eval\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: intermediate\nTarget language: Bash\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nEval:\npass@k, time-to-green, regressions, diff size\n", "reasoning_steps": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "ci_integration", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec00dfd4fc667ead8bd61bfed10ef77f5dc0e37a"} +{"id": "train_09990", "created": "2026-01-01", "topic": "Dataset curation pipelines (filter, dedupe, quality)", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Dataset curation pipelines (filter, dedupe, quality)\nDifficulty: advanced\nTarget language: Go\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fab3f21cf7eac765b84bdcfa3fbc934032729cb"} +{"id": "train_09991", "created": "2026-01-01", "topic": "Model merging, distillation, and continued pretraining", "task_type": "code", "difficulty": "advanced", "instruction": "Task: code\nTopic: Model merging, distillation, and continued pretraining\nDifficulty: advanced\nTarget language: TypeScript\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65b60ebd808ca65d492450c645e77d0b0946a080"} +{"id": "train_09992", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "review", "difficulty": "expert", "instruction": "Task: review\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: SQL\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nReview: correctness, security, performance, governance\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f684fa80f91d2e8d5463370fc187d9d9fb8b7a4"} +{"id": "train_09993", "created": "2026-01-01", "topic": "Self-improving agents and feedback loops", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Self-improving agents and feedback loops\nDifficulty: advanced\nTarget language: JavaScript\nContext: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "cost_latency_tradeoffs", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e2e8661445c6a6b5174a7ef028118e23a5805ad"} +{"id": "train_09994", "created": "2026-01-01", "topic": "Tool calling, sandboxes, and CI integration", "task_type": "patch_diff", "difficulty": "intermediate", "instruction": "Task: patch_diff\nTopic: Tool calling, sandboxes, and CI integration\nDifficulty: intermediate\nTarget language: SQL\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db3a13583c012f22be5a9016a5e7b5815f95bba3"} +{"id": "train_09995", "created": "2026-01-01", "topic": "Multimodal dev workflows (docs, diagrams, traces)", "task_type": "failure_analysis", "difficulty": "expert", "instruction": "Task: failure_analysis\nTopic: Multimodal dev workflows (docs, diagrams, traces)\nDifficulty: expert\nTarget language: C#\nContext: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\n- Initial patch broke edge case\nReflection:\n- Missing zero-input guard\nCorrection:\n- Add explicit validation + test\n", "reasoning_steps": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "documentation", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c891defad4c09170056b2348d6fbb6b80616f5e"} +{"id": "train_09996", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "patch_diff", "difficulty": "expert", "instruction": "Task: patch_diff\nTopic: Secure code generation and policy gates\nDifficulty: expert\nTarget language: TypeScript\nContext: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ if x == 0:\n+ raise ValueError('division by zero')\n```\nAcceptance:\n- Tests pass\n- No new regressions\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "tests_are_truth", "governance", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d57bbb9d739df45d90915c9be145975f2767c3c7"} +{"id": "train_09997", "created": "2026-01-01", "topic": "Secure code generation and policy gates", "task_type": "design", "difficulty": "advanced", "instruction": "Task: design\nTopic: Secure code generation and policy gates\nDifficulty: advanced\nTarget language: Go\nContext: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "auditability", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f3628b4bc4a9634bfc7f92c88fa8b482b2105f1"} +{"id": "train_09998", "created": "2026-01-01", "topic": "Reasoning-first coding models and tunable deliberation", "task_type": "explain", "difficulty": "expert", "instruction": "Task: explain\nTopic: Reasoning-first coding models and tunable deliberation\nDifficulty: expert\nTarget language: C#\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bc9ed12ff71c364809defe21a623a16abef67c9"} +{"id": "train_09999", "created": "2026-01-01", "topic": "Governance, provenance, and licensing for code data", "task_type": "code", "difficulty": "expert", "instruction": "Task: code\nTopic: Governance, provenance, and licensing for code data\nDifficulty: expert\nTarget language: JavaScript\nContext: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance criteria\n", "reasoning_steps": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "documentation", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3947672bba12a98562d5b63de2e16590cd462746"}