diff --git "a/data/train-00008.jsonl" "b/data/train-00008.jsonl" new file mode 100644--- /dev/null +++ "b/data/train-00008.jsonl" @@ -0,0 +1,5000 @@ +{"id": "train_40000", "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: 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": ["auditability", "documentation", "governance", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d607d36c02fa07605d620667e756a70b46945f9"} +{"id": "train_40001", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f05fed70d41c575b9ea166b634b01b8028f56d75"} +{"id": "train_40002", "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: 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": ["tooling", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06e196282aa7159325706d4ea776afebe433ced9"} +{"id": "train_40003", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["documentation", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0a9592f90f3d58b830f6dcc357b867eb9381b66"} +{"id": "train_40004", "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: 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", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2161c43bcf36eef238811cb92523725e89566e8"} +{"id": "train_40005", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5422f429f89f5c2278e51618e33dee8d39f82ea"} +{"id": "train_40006", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37fbabbba2fbfa5276b58e25021bc8a3da22b358"} +{"id": "train_40007", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec4d0ef8e472999a6002a635706aef100ae9d772"} +{"id": "train_40008", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9780a9794513b467e02aa370bd65509609c6cbd7"} +{"id": "train_40009", "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: 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", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcdad556f625a8755df8080ff4def0e95d1be9b4"} +{"id": "train_40010", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "documentation", "security_gates", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fb53b304994a1e7327b4d6c9d0b403e161190fc"} +{"id": "train_40011", "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: 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": ["repo_scale_reasoning", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1a57b97529ff0aa96e152049b448d02198cdba6"} +{"id": "train_40012", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["tooling", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bff923dd4f65887cf72856ba328e9b2df3e3af99"} +{"id": "train_40013", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2dcc7dd4e77e13279ae51e82201a3d52d7f01cf9"} +{"id": "train_40014", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3929bbc82635c888aed50c2d25b810c85c03acbd"} +{"id": "train_40015", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f133a1e823c67a6d9c68aa7672908f524fa55c73"} +{"id": "train_40016", "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: 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": ["cost_latency_tradeoffs", "reproducibility", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19d6f2ffca5bff6e94e13122aef0679079a5da3d"} +{"id": "train_40017", "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": ["cost_latency_tradeoffs", "governance", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86d13ef34d55cf2d9e1c0420ca7718dbfa1b0b58"} +{"id": "train_40018", "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: 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afaad93acbc3cb94b230f3922f43c6c58d4bf3aa"} +{"id": "train_40019", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b5188e1c26b9beee427baaf79b5ae5d9c702b81"} +{"id": "train_40020", "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": ["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", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca7a4a3c04b26b713e6569c3335da9ad3cf8404e"} +{"id": "train_40021", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["documentation", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d59488dd0ff09161b856cf2c1d49db296d7ed73f"} +{"id": "train_40022", "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: 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": ["tooling", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de050614581577988242c078a91ab61c0208deab"} +{"id": "train_40023", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "governance", "ci_integration", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "176f48651a1b27835a5ef9c55aa4bef5a2c0f638"} +{"id": "train_40024", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16f6e237461e76149fb67a9214e2b017edb309e3"} +{"id": "train_40025", "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: 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d4cb85cabe58bd8c179b0c45c4e91e9914086c0"} +{"id": "train_40026", "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": ["tooling", "auditability", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "985aff40728fac272cb3c476278c1b786c793194"} +{"id": "train_40027", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea3ac5b3194d782f4dd229889537b9a50771175f"} +{"id": "train_40028", "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", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0df682007ef6a32f480d90e8b889d6231c3e8fb"} +{"id": "train_40029", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "documentation", "reproducibility", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c62e4a27f2e8ce6aafd39b3452bfd97ef65fcd96"} +{"id": "train_40030", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready 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", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc7a8454d4bfe4080131ac1a52ea78e9c0bf3f4a"} +{"id": "train_40031", "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": ["governance", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52a634069a9799c4d7d2b247f6385543fa6c5e68"} +{"id": "train_40032", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2400b742dbdee57a2fb4c4be64ce9e14f0ecaf9"} +{"id": "train_40033", "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: 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": ["ci_integration", "auditability", "tooling", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "448ae4ac8fd93f403e29d603c0ac27ad22d7bfe2"} +{"id": "train_40034", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb6a0f99f3c0344d9ae461caa295c0ce2069127a"} +{"id": "train_40035", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0fe7b5f45bdff38c69e575b9323781cba64ed2e"} +{"id": "train_40036", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "auditability", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92dafdc981b5b62a2da95cf85e583566a9221922"} +{"id": "train_40037", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cd81a5b60b59ddc08cc99e2cd9778e8a9ea8442"} +{"id": "train_40038", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12f4ce398d532d76d48872641e624eefa42f35df"} +{"id": "train_40039", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24a65d412d27a936dcffa0678c7a0249bd92d64f"} +{"id": "train_40040", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e152ad1ced1cb76fb3a24a9cbb7248b70ae6c06a"} +{"id": "train_40041", "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: 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": ["tooling", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10f5b24fe4e43d5823abf3f9d7dd1eaadbceffcb"} +{"id": "train_40042", "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", "auditability", "tooling", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e66ab97d6ce31b7464d932aa189f1c980bb3c59d"} +{"id": "train_40043", "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: 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", "security_gates", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60a580e05487b5048eb20072dd209692d7054528"} +{"id": "train_40044", "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: 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": ["auditability", "evaluation_metrics", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "408942538dc6fa7f8c54adc5e777c5b9348983e2"} +{"id": "train_40045", "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": ["tooling", "evaluation_metrics", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "368bce8a3298bb32ffb36e1110bb7e5f5f63ca0e"} +{"id": "train_40046", "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": ["auditability", "security_gates", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d614ba08a0f36713fd13e7fb3b5820e7a854f80a"} +{"id": "train_40047", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6027949719b1ac6293ed2a1657a0fe5a9c5aba96"} +{"id": "train_40048", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83651bd644a0d4faeb46f9741b8e0c9d53c274cb"} +{"id": "train_40049", "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: 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": ["governance", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1fe1247437369d86eb7a1574498bfacd4654a05"} +{"id": "train_40050", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73e0bc61a90f00e4635b2dbb5e974eb9efb5bf56"} +{"id": "train_40051", "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: 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": ["evaluation_metrics", "documentation", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05949bdfe468fdbdd8a837e68d5d7e9f0ded2f46"} +{"id": "train_40052", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "cost_latency_tradeoffs", "security_gates", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73d53c9590ee4deeac43dc281a4d088c8effcf82"} +{"id": "train_40053", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "ci_integration", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f31a77224ddde94109b913933f07a53eb67e0ee2"} +{"id": "train_40054", "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": ["repo_scale_reasoning", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f3628b4bc4a9634bfc7f92c88fa8b482b2105f1"} +{"id": "train_40055", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "tooling", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd44d90360a3ce318d228b25e28000df4917f49a"} +{"id": "train_40056", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["security_gates", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5cd75448bc9feda928a48a644b7b7ebb9ecc8621"} +{"id": "train_40057", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tooling", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c3b543b060ee641ee1a56dfba874bbd8fae471e"} +{"id": "train_40058", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b353c2913e5d701928871a8bd4cb8a1a6c4ef16f"} +{"id": "train_40059", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5aa5fb916c48fdfca5bd44f089b164fd67c984cb"} +{"id": "train_40060", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5030aa1b9726ab67f8c3190f07df482179a68f3f"} +{"id": "train_40061", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62df666d82a9df7dd3d5541e44446865a33cb326"} +{"id": "train_40062", "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": ["ci_integration", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0df9c4aab35f71573e570319036f65518fcb3ddb"} +{"id": "train_40063", "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: 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": ["governance", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e92a7cd52793efab20e35c2ee2f74ca68820e0e"} +{"id": "train_40064", "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": ["tooling", "ci_integration", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a63cdd152c163f5905a8258f3889e176889332a"} +{"id": "train_40065", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2473f51694784fd2f1185fe40aea3764ebf5d0cb"} +{"id": "train_40066", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "349aabb280825d80a696f539cac48692209a6447"} +{"id": "train_40067", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df4b422910865bb28b75a3dc302ca27e3ebbef3c"} +{"id": "train_40068", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6ba995b99e0b73a48beb11d6c090ed0f614a5ca"} +{"id": "train_40069", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f78201193d3c9428c67bac795b5a039a76add2d"} +{"id": "train_40070", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e9e99679dc32584ae0605fe023e63d52bffcd01"} +{"id": "train_40071", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "evaluation_metrics", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc9b760ce9b5d494410944a43474446959b2156f"} +{"id": "train_40072", "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: 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", "evaluation_metrics", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61b6e02a964add1f7e74041b4306e44e817d4e39"} +{"id": "train_40073", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50aff152a0bfb350b4f9a5db185efb931b0b91e9"} +{"id": "train_40074", "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": ["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", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df0a95a24222c2c41295c413f5712b5f4531f120"} +{"id": "train_40075", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "deccff884c62fcf0df8c6ba9bfb33d5be1cd287d"} +{"id": "train_40076", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "ci_integration", "repo_scale_reasoning", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8657d025ff2fe6778a2e5c4f86f5e6912240bb9b"} +{"id": "train_40077", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fa8357ec1ab1021b58c5b0aa4384204ff2fa7ec"} +{"id": "train_40078", "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": ["tests_are_truth", "cost_latency_tradeoffs", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "767ac1923594a1f790bdc7cec8a45d5c39dab5aa"} +{"id": "train_40079", "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: 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", "governance", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d42162a837ea62d0db6a8b2be46cd6433535f099"} +{"id": "train_40080", "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: 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": ["auditability", "governance", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cfd3a085c6747710841efe0b4d62809c8f260b6"} +{"id": "train_40081", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c275adc0c5e6747b88dad1e41bd60f22b01560a"} +{"id": "train_40082", "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: 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", "auditability", "governance", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c4ccdfebc7360c6033f20a48ce0080d01c968bd"} +{"id": "train_40083", "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: 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": ["tooling", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6026552670784474e60e9bc49f59e65cbb6783c"} +{"id": "train_40084", "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": ["security_gates", "tests_are_truth", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca5e5447cbbdf790849d38141c4ff59f07e5756c"} +{"id": "train_40085", "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: 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": ["security_gates", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba57200bcd1261b0a4c984f0aa18489642c4cae7"} +{"id": "train_40086", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d8ead2c234b5cdcde3512e42cdf0e93b9c32cb5"} +{"id": "train_40087", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["reproducibility", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a45530e15ab06e7db892f5e097c3cd53beb307d3"} +{"id": "train_40088", "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: 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": ["auditability", "ci_integration", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "268e41bc8fbb5397824dc23c175066f15cc2b98c"} +{"id": "train_40089", "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": ["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", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2531974312a81dbe7f6475d873732f4ff049349"} +{"id": "train_40090", "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: 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": ["ci_integration", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4052750e833a358f2137d142f339c76ffaf8f560"} +{"id": "train_40091", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f37c48ba3eeb92b8f9f5aca47310930819edafec"} +{"id": "train_40092", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69ec6251d327eac16e427befb43542a50d30ef4e"} +{"id": "train_40093", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c29c84e0ef5abd3cb95dfb4a5512579d5e60b9c2"} +{"id": "train_40094", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e26b7ad0ab5b0c841c3b81ccb22c0e46b6370240"} +{"id": "train_40095", "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": ["governance", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0cb83836b70a249b4581176e3868d405ef755b5"} +{"id": "train_40096", "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": ["cost_latency_tradeoffs", "tooling", "governance", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2bb68bb34b6749ffdd4922fa2d6c577d4f13d27"} +{"id": "train_40097", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f98116376953e41892e75f12eee09cbb87ab348"} +{"id": "train_40098", "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: 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", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb914068491bb03d38d3d698c8943246af51042c"} +{"id": "train_40099", "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": ["ci_integration", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8c28525570c3b36ec029370ca1e5d2120fc8219"} +{"id": "train_40100", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7244ef29260d2dbb7f876dcf83a44e5b593cef55"} +{"id": "train_40101", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdc04a943bb0609c09ca591d215ef52a44234a27"} +{"id": "train_40102", "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": ["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", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca8c968dbe7e5d50af8145879e2a3d7c856cd161"} +{"id": "train_40103", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["tooling", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85d38e0f51a7dbfc07e874f4553e03dc65e8bb44"} +{"id": "train_40104", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47fdcb15f82b98d7ddb1032e20c82b5e12260fa0"} +{"id": "train_40105", "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": ["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", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b013c9b886c5fd90f4a9efd1b0f7ad88fcc51de3"} +{"id": "train_40106", "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": ["tooling", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e72dfed53629201d4721cb086a85ae71be4e2493"} +{"id": "train_40107", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["repo_scale_reasoning", "security_gates", "auditability", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b283cf8c4df88fa594e62868da45c6bb6f3da24f"} +{"id": "train_40108", "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": ["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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12886fc7bcb6afc74fdc47092ee07e4c853dd4ca"} +{"id": "train_40109", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fb16dd2ae8fc8333ae10570362e89b0c7962122"} +{"id": "train_40110", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2dde083a93a0bc415bfbde90b225214adf0086a9"} +{"id": "train_40111", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd3353fd2a27f5b60ef8a6074b00007c906c6530"} +{"id": "train_40112", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fae7d1923a92aad0368fedad0c6154251bdd5a89"} +{"id": "train_40113", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e68a0da6665682b104a06fdfb380eaea0dcfa6da"} +{"id": "train_40114", "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: 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": ["security_gates", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05f7929fc9bdfb01614f87efc5a937282abc4e2b"} +{"id": "train_40115", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d8c147951f59ddd206404bd8749f7db297c2fa8"} +{"id": "train_40116", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dc03343d0f1c64d22312c4691be9d3c3a228194"} +{"id": "train_40117", "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: 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", "ci_integration", "security_gates", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06819934213fd5603778137f5b6c85d7bdcab442"} +{"id": "train_40118", "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: 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": ["tooling", "auditability", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af7ef924754d390d7d81e6b2deb280e224347b3c"} +{"id": "train_40119", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "documentation", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f14aa7049ef3c204b254ef1445e691912c401a0d"} +{"id": "train_40120", "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: 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": ["tests_are_truth", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ad91af11ed56d7fdef164c0fd9fa1734f66f89d"} +{"id": "train_40121", "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: 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": ["tests_are_truth", "reproducibility", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b788162b3f1bfceb92c177943f6a67439824a2b6"} +{"id": "train_40122", "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: 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", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e1c26750c2a0818cf399aeaea917e7a5b84cf62"} +{"id": "train_40123", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff578cc433b1d65ac86d1c703a0e925f7f3bce37"} +{"id": "train_40124", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4782e2c2391103d0ac31cdef9edd2a9addda4e5"} +{"id": "train_40125", "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: 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": ["repo_scale_reasoning", "ci_integration", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "860faa426b4c7e546964cc0c7e7e8ade0bc9deb0"} +{"id": "train_40126", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fa36afeb62e9deddd8cccf36ba0def7f317576c"} +{"id": "train_40127", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77461fd72bb73d847f12599e19be50f4a6da4444"} +{"id": "train_40128", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "tooling", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b49208bc61a8c5f7085ab5118fcbadce4c085be7"} +{"id": "train_40129", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e10ca851f24eb3468bb0eb8fd2903538741b7cf"} +{"id": "train_40130", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a51c9dd11d5ef6f9ccafc46ff751169affb39dd2"} +{"id": "train_40131", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "repo_scale_reasoning", "tooling", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d269ca425d333a8f33a8de5d98f665153caa1410"} +{"id": "train_40132", "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: 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", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f7d58ab5c7f67a3642d2753dc8b140d5a881522"} +{"id": "train_40133", "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": ["evaluation_metrics", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3543ea9e9b28cace873220a9d326c35e7d8711f2"} +{"id": "train_40134", "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": ["auditability", "documentation", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cf57616840bea41181ed955f5b000ba94770239"} +{"id": "train_40135", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4875fe79c325c17495988b9fc9a95cea7cba4ba"} +{"id": "train_40136", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fddfee7f7af802ac1b6547f53182cc59cd0329f9"} +{"id": "train_40137", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25d5f1800edcb411b4e5a40a88519779d71b7cda"} +{"id": "train_40138", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["security_gates", "tests_are_truth", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e8c0772a9436d01ce445490e1675a82d1d14dfa"} +{"id": "train_40139", "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": ["tests_are_truth", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23502e1f6203d8167ded3e2d25b3a73a0c4b7277"} +{"id": "train_40140", "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: 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": ["security_gates", "ci_integration", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0652961542fda71492945707dbe0a4b5aff089ae"} +{"id": "train_40141", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf07be9916e9be1fd4ee8d34ded900b0954c25af"} +{"id": "train_40142", "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: 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", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54942d9c9bc0ecbbba042d28bc61b74dae0ac974"} +{"id": "train_40143", "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: 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", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0775307aba77d57e4ba1da647e9e4ecf7faa07ec"} +{"id": "train_40144", "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": ["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", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4ae3f986859324ee985953126efc351f7b76aa6"} +{"id": "train_40145", "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: 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": ["auditability", "documentation", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e8338a247da74c90fb31648746a2f551bccbf25"} +{"id": "train_40146", "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: 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", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "764de3cc79f97f860700051d63420e369529e88f"} +{"id": "train_40147", "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": ["auditability", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc2d5a3c63c6f5ad0af2edf3269a5282ff6a5800"} +{"id": "train_40148", "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": ["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", "documentation", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c75c543333e41b4a30f7ff4555e4c468dca3bfe"} +{"id": "train_40149", "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: 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": ["tests_are_truth", "evaluation_metrics", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9db348de33829b93454d2dcdb089a1adb18eb3ee"} +{"id": "train_40150", "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: 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": ["tooling", "tests_are_truth", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2de44013b2ddf26479bb47b56c2aa01b4894cba5"} +{"id": "train_40151", "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: 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", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84744ec0981db3f23e11769e0469cf195f4c04f7"} +{"id": "train_40152", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a57904159b1f5d447aed044e843c744128ea2fb"} +{"id": "train_40153", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4912312a2dce020b4c1d6871f64edb53719f63a"} +{"id": "train_40154", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf184903034db28f83f705803bb440897e823c3c"} +{"id": "train_40155", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56598f526666cb03f3148b74a8e20b353b66d3e6"} +{"id": "train_40156", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26bb7349ab6e0f3b7872cad5869d1565c14454d7"} +{"id": "train_40157", "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: 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", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdaa9ccd9ba9cd7c7efbf9eebe4af3a7212a771c"} +{"id": "train_40158", "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: 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": ["repo_scale_reasoning", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f371044665103d7851bfbce6581fe856bde29fc7"} +{"id": "train_40159", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "governance", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d494d2169b295f6dbf25c370c3e9ede534e9e5dc"} +{"id": "train_40160", "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: 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": ["documentation", "ci_integration", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f3357e04defac47f5ced30fd6a8184e0ff6265d"} +{"id": "train_40161", "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: 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": ["documentation", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecf2d406081d83b984b37c361d2a8e4f388f6112"} +{"id": "train_40162", "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: 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": ["cost_latency_tradeoffs", "reproducibility", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f222e54c10d7c498cfc42817ed0bfd8f594989bb"} +{"id": "train_40163", "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": ["ci_integration", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cab07b1db1382b0a1b75c5a9517db92e3b5fc59"} +{"id": "train_40164", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["security_gates", "reproducibility", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1330e0ac71028d4d99e8b72ca4f2518d12394158"} +{"id": "train_40165", "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: 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", "reproducibility", "security_gates", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af9b7a7f3a45e896b966af5b370bd690f7150b26"} +{"id": "train_40166", "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": ["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", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc512f9d77d213bcc84f8854dc3d4082371054ae"} +{"id": "train_40167", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ac44521dcf0f3a9aba75be61735488bc59a1494"} +{"id": "train_40168", "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: 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": ["security_gates", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "421f67df90eb0ebfaf66e9429038a959963d1d7d"} +{"id": "train_40169", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d9fbeb976f2298cc106ab7f46f107c716bd8d1c"} +{"id": "train_40170", "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: 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", "tooling", "auditability", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "910dfba72454eed6056b37c85265d0ee53cc5eb7"} +{"id": "train_40171", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready 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", "documentation", "security_gates", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35b34ba03008105a4644eeddaad627d795d26eb0"} +{"id": "train_40172", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff5ab38f25ef15a36a0765aec275edda38dff370"} +{"id": "train_40173", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c44fb00b2fd77c2020b7097fcc440a310956456"} +{"id": "train_40174", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14317ab0a64f1cea593df5b10d539421cf71826e"} +{"id": "train_40175", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0486f96bc1b423d0410a3ce4d1f4609da18ee3c9"} +{"id": "train_40176", "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: 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", "documentation", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6719b931bc24cbdbb184c2d2209c4af014f979a"} +{"id": "train_40177", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["documentation", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2669d0e5763b5ed41a8640b5eda1a49d5e19a854"} +{"id": "train_40178", "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: 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": ["tests_are_truth", "documentation", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64d51daedd84faec68ba3ad512967d3daef2bcee"} +{"id": "train_40179", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0ebe1699ba47bc53a6f730a2663fb609ceee97b"} +{"id": "train_40180", "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: 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", "tooling", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3afbf6e988d3ac7b3b1f9203d4afc08b3faa29a8"} +{"id": "train_40181", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tests_are_truth", "ci_integration", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d399062816d5d938772e7406cf5c6c47650b61c"} +{"id": "train_40182", "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: 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": ["reproducibility", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d83d72f9920168f02b85c4b9e62c74d7086ef21d"} +{"id": "train_40183", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["ci_integration", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c9a37c1a44c4bc826b7164022c96529ba6189f9"} +{"id": "train_40184", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4137cfba47f222c44ff04a077073b8bce8610bd"} +{"id": "train_40185", "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": ["governance", "cost_latency_tradeoffs", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9de37f70427580b12ea23b7907404f77c4c89465"} +{"id": "train_40186", "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": ["auditability", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d86d38ef2992abaa6e78359d0475089870cecf97"} +{"id": "train_40187", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16d040084b72b37b8ef8346d68f13ab7ba27fffe"} +{"id": "train_40188", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85763caac51b7d2b044b0668d8158dcd86384e07"} +{"id": "train_40189", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d2e203c10288c2992cbc29de070acfaef77e0d1"} +{"id": "train_40190", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6e4b68b1daec963bac5c8c77f69c6e7fa0da41f"} +{"id": "train_40191", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "security_gates", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca4f2386e34b94c8d394c000f9b2afa8d2316bb3"} +{"id": "train_40192", "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: 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", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d35d8613b94549bc42d69ce877b60d6728c0dfb"} +{"id": "train_40193", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccaf94f2fdd1a6363bf330791e661ef154643259"} +{"id": "train_40194", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52dfac0615b9ea51c8943c153a5690e144a56f8c"} +{"id": "train_40195", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "698253205af94eeb763b5276ef4021507eb0ab09"} +{"id": "train_40196", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd2ca39e588e9a7aa27fc2268acf6692963de60c"} +{"id": "train_40197", "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": ["tests_are_truth", "governance", "tooling", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55603140465e5cfc35176a0044ca828a25619e18"} +{"id": "train_40198", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8984656bb921cc2ab16352328f2869e7c2eeb71f"} +{"id": "train_40199", "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": ["ci_integration", "cost_latency_tradeoffs", "documentation", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfa9ad1db229dc876b5430bc17b4ecd6cd44b379"} +{"id": "train_40200", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b090e561f9b6ff8a7ab9c2a35e335d7e211d2350"} +{"id": "train_40201", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8494c0b9a04cc1e6f3b595133e46182b8b18724"} +{"id": "train_40202", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["tests_are_truth", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92994400e6e5d57b097b963e60cbb9a811ecd520"} +{"id": "train_40203", "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", "auditability", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18b8a6cd6e976e370749dab790005ddfdb0fc6fb"} +{"id": "train_40204", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef00630c8239c4af69078fffa7e8a74129a330f"} +{"id": "train_40205", "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: 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": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5cb7c1c54d56d710154c2ab25afd23a45049ef0"} +{"id": "train_40206", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01c101858cb3cc1b6965d52c607490556b671b2b"} +{"id": "train_40207", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25ef4592b9588f687e3530fcbecaf0857f7988fe"} +{"id": "train_40208", "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: 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", "repo_scale_reasoning", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ac85372470a841d0c0bd057926036aa08f8d5ac"} +{"id": "train_40209", "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: 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", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1c29c042f8e4ffb916fdf88d2b4403acc9ea9ff"} +{"id": "train_40210", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b81edeeece53dbaa70c82537574207c17bd6b6f"} +{"id": "train_40211", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99e8bdef1b580e37bb65ab670d9fa174454e2dc9"} +{"id": "train_40212", "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: 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", "reproducibility", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f657141532ecdd7d29c17a340eb7b27bbbd0552"} +{"id": "train_40213", "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", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62bdb40fb861a019bee08c2564234fa5e3dfdd0a"} +{"id": "train_40214", "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": ["ci_integration", "tests_are_truth", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "149e173e194bd55408d1444ce9690b9c96e1a483"} +{"id": "train_40215", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15661ae8085f8278df98a17547507cc30a2db2b5"} +{"id": "train_40216", "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": ["auditability", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2169545f652da9c91dae2cfe11d50ee0fba66ac0"} +{"id": "train_40217", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "repo_scale_reasoning", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7581fb642a66aea84b3e47ff12f065277a1fbf29"} +{"id": "train_40218", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "035d4df7a0d1380b9ed3a5abaa3b9800149b5aef"} +{"id": "train_40219", "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": ["reproducibility", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a200c501fb823f68d1300f2ef1219f5fcfd4b9e"} +{"id": "train_40220", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7c17d55ffa775820e102a0e5e59ac3899ed0b94"} +{"id": "train_40221", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1323900fd0d2d5ab37ab2b80eff46d315d6ded76"} +{"id": "train_40222", "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: 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": ["ci_integration", "tooling", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2689b65890e287481adde83922e73aaeef7f95ba"} +{"id": "train_40223", "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: 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", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3119d0afdd4002b6ad8bbedb34605dbcda64cff"} +{"id": "train_40224", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9ca796c5312ebbdf2b2854a63ed6cb7717433ad"} +{"id": "train_40225", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5601366bfdf46bdd7c9fbca11c8e4200aba0616"} +{"id": "train_40226", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7b9b45fbe75c947c8c426b221e25ce683796355"} +{"id": "train_40227", "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: 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", "security_gates", "governance", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f6bf9a5bf4c2a34f8fb2bf3e00cd3c168f08b9b"} +{"id": "train_40228", "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: 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": ["repo_scale_reasoning", "ci_integration", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccfd379fb189cf7663b1aba7a919f33f9df4d827"} +{"id": "train_40229", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b30019b198a402048bb7d949a3ded373acf4844e"} +{"id": "train_40230", "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: 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", "auditability", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2aa80d41ac45bded7f937cf4b6768bc02686252"} +{"id": "train_40231", "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: 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", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e525a6547218ec6f8f2752215b3def8a281da28"} +{"id": "train_40232", "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: 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", "ci_integration", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "612be51a66e8ba7a7aefcaf1162cbbe1969f8cd8"} +{"id": "train_40233", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d55dd5e875f19e9027d27f054b9333bb985309ef"} +{"id": "train_40234", "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: 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", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "acbdb731f5dcbd636f3e44e8c6cc65dc49edf9cf"} +{"id": "train_40235", "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", "cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4de82b71fc04630712bc98bbe131eceedcd378a0"} +{"id": "train_40236", "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: 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": ["reproducibility", "documentation", "security_gates", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31f13250e4d97545747dbdee1ab9a99a0fce4c14"} +{"id": "train_40237", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5559d4c2abbbb0cb86805b885df0e9e5e4fd82f3"} +{"id": "train_40238", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09244356bd51934054635565db0031b08c495afe"} +{"id": "train_40239", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "866254a7dac49c99f9fe7da7d7b66ff5666f84e3"} +{"id": "train_40240", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ab2bab09b698268c953007277d4a3111ffd04d2"} +{"id": "train_40241", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42b0f395687d54b1d49afcffa8f6e3583c61bac5"} +{"id": "train_40242", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33c4f6eb3bfaa3a9ef50dee9ed6a8f9d876554e3"} +{"id": "train_40243", "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: 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", "evaluation_metrics", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdbf4e7c36b32f6689ec7272a4a4794a16132f57"} +{"id": "train_40244", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "documentation", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e8df779c61477f42de714a26df0a32ff7c54981"} +{"id": "train_40245", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27b33fc26bdfa8fe61a4f7fdeec0efe63088c249"} +{"id": "train_40246", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e336cb787159ed21c72c7cf9e52e1ccacecba5ad"} +{"id": "train_40247", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c43991833904ac1ea13c66ec61447394ac85499f"} +{"id": "train_40248", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "governance", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cac8f80667d1b8b067ab89fe1eac93042af981b"} +{"id": "train_40249", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3598eb91445c31ad563aa91e2cd49d4dc81ae648"} +{"id": "train_40250", "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": ["security_gates", "governance", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7eeb2ec63c426fcd69cf73bd0ca9fbfe760e4755"} +{"id": "train_40251", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ce51f9cae436ca38998375dc96cc47fffc5cf16"} +{"id": "train_40252", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "reproducibility", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8c2e95512c58bd86cec224cc5bdd7ef70354767"} +{"id": "train_40253", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready 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", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c01649f523a454b1ea35b9260c38dbea3ff3dd52"} +{"id": "train_40254", "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": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8bcdedd9d202d4e02525aa497d45fc2f9d80b57"} +{"id": "train_40255", "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: 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": ["reproducibility", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f740b259d0cbf946d86b10f1bd805c985336926"} +{"id": "train_40256", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e197d890c5c71e04387288e1f9b6c1cca7bdf39"} +{"id": "train_40257", "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": ["cost_latency_tradeoffs", "evaluation_metrics", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31d961f4d27ccfed4a2b193f59f2747d1dbe8216"} +{"id": "train_40258", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd85c302046ba83c4bf6d93ae6d186c1383f9275"} +{"id": "train_40259", "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: 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", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b350fb02a991a7afe0eed1544853c82fdb68d49f"} +{"id": "train_40260", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b8317afcadca7b31857b91f7061795e4d0c8cae"} +{"id": "train_40261", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b4aff3c737bea56fa142630b99d22ebdcf58f12"} +{"id": "train_40262", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e0d0cad165c2a9644945844a3f44f817fc43da3"} +{"id": "train_40263", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d59959c9669f2d5fd7fd4fbeb123400033a27af0"} +{"id": "train_40264", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "863363bb0116dc036835424f8a9f51036b513fdb"} +{"id": "train_40265", "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: 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": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d01165a3fa8cae06f1ba3a50a54c9323dd027042"} +{"id": "train_40266", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08a121f87f85a2f0658eb1d31d52de83ce6312ab"} +{"id": "train_40267", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b37bc8ae873dad27500be6b23317f466e8c0263"} +{"id": "train_40268", "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: 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", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aac70e3742d011cc301a589130f989e698dbdc45"} +{"id": "train_40269", "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": ["tooling", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ba069f0a55e73fe462b825ba88bc6094404266e"} +{"id": "train_40270", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "security_gates", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64b8049780989337f15324b95064371b31529467"} +{"id": "train_40271", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8da69c0a4048771139e1ed9de582af0d623ff43"} +{"id": "train_40272", "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": ["security_gates", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "313b50d1a34199cae2af08f1bb58f143e7eeff6d"} +{"id": "train_40273", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1cd5b6fd3562917f950a8bf8d68a356078c90cd"} +{"id": "train_40274", "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: 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", "security_gates", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45fe5ec34e482ec54c8d8a2e6a5c8a261f9b4438"} +{"id": "train_40275", "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": ["evaluation_metrics", "governance", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0aa85222695fd7c94896f8461be7e4576070456"} +{"id": "train_40276", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "148ebcc3a074349e4e505bc3b92f9d7c1df1861b"} +{"id": "train_40277", "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": ["security_gates", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d89fd6342392d5e9d9853618ea25c4bfe695139b"} +{"id": "train_40278", "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": ["cost_latency_tradeoffs", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f3f9253c1294fc84792549d601021af01ef1893"} +{"id": "train_40279", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "repo_scale_reasoning", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b71ac3dc51c8ddeba5d275bb175daf16492047bc"} +{"id": "train_40280", "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": ["documentation", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a1b381b889cef02cd7b096d4ad8d589d51011b6"} +{"id": "train_40281", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d77c5452c681cc56f96b8def880aeeac9543235"} +{"id": "train_40282", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96e6a9d6f0eb6490d6c0d8bf4d38bcd89130923d"} +{"id": "train_40283", "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: 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": ["repo_scale_reasoning", "governance", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1533cee1db8a38c6d5e3aadd2ed4d6f346a6dfba"} +{"id": "train_40284", "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: 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": ["tests_are_truth", "tooling", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71cad1fe0867f9bfc61c1b9a41c040f3844c3647"} +{"id": "train_40285", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0104c6e077b49091f0b492592ab6f86d133d4c7"} +{"id": "train_40286", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "security_gates", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79a89ba0ac8ca778d0edae9d72c243654d9f2096"} +{"id": "train_40287", "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": ["ci_integration", "evaluation_metrics", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e494702397a5b3200556ad7abfd14d439dcb960"} +{"id": "train_40288", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "027729bf54b1a62388bf6b395b13a94dc658455b"} +{"id": "train_40289", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "454fdcb2ae3bd1dc389ffba57f5480c4588ae84e"} +{"id": "train_40290", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "evaluation_metrics", "ci_integration", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "611c681549d82a0f0fe822c856f32c2bfd599f80"} +{"id": "train_40291", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1558f4e2afa32fd0ef5c767a02ea7f20a5115159"} +{"id": "train_40292", "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: 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", "auditability", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5b4ede8b08424002a7aad29eaf31295d9cbb16b"} +{"id": "train_40293", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e5076117d6c9178a8692e80bbd8a4de401e5a2c"} +{"id": "train_40294", "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": ["tests_are_truth", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d160c81e40bb990a876632aa751860372af063b7"} +{"id": "train_40295", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f939bb0f7c4cc66c7a2cd4faab8bf59f4960d9b9"} +{"id": "train_40296", "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: 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", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9dfecbbb89f0ff5bc6a212a996e9525fd9b5beca"} +{"id": "train_40297", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e6898bceec258d36069737257012d3f3f1ba8de"} +{"id": "train_40298", "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: 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": ["reproducibility", "tests_are_truth", "tooling", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "270a22ce73764aa022faf6e36ef9c4fc893e3bb8"} +{"id": "train_40299", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "889a51edcb34e0656885ea0451eec0f6343c0b42"} +{"id": "train_40300", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd64113a63386e5631a7640610b3ac77f1817cdb"} +{"id": "train_40301", "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: 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", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f3044bff0682498280b7fe61609125b3c556d00"} +{"id": "train_40302", "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: 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": ["tests_are_truth", "tooling", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ba751efd2cc7ef6c3ee460e437bf13b6b5f8656"} +{"id": "train_40303", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ceb9dea717a254738fa5ba9365dcb15e533ddd7c"} +{"id": "train_40304", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b32c04cac2b50e850a21cb1913aa8a8fa18450f9"} +{"id": "train_40305", "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", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa545fff4caa8a38ced4b1d09b856157a4f68de3"} +{"id": "train_40306", "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: 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", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "449e8aa7daffdcd92c4a729c9dc89643eab1cd0c"} +{"id": "train_40307", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c91c66799b2a2aec58dd716bbbe798ae74b8ea9c"} +{"id": "train_40308", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc4d75ef0267b893b9a67b60ffc642475e9de2eb"} +{"id": "train_40309", "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: 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": ["repo_scale_reasoning", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6309203e240aec3a3648bdb0f9a5f44fe7d3cf0"} +{"id": "train_40310", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cd17ce8bdd59cbb92be5647e8a8107b10ec54ca"} +{"id": "train_40311", "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": ["security_gates", "auditability", "tooling", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "075463ebe7e3768c7bc861795b3adbef6227a52f"} +{"id": "train_40312", "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: 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", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82e080a94698db2f56d7ebb9390b521c64335393"} +{"id": "train_40313", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fda354c3daa29a5f7812c0a41b7cea04627502dd"} +{"id": "train_40314", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "022dc1f80255788a950a7dc202c457914449cb64"} +{"id": "train_40315", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0717bbab8d779be708723acbad2e7ce1bdc87327"} +{"id": "train_40316", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1ffbcfccd994f02dcddb55ea478fa91bf310587"} +{"id": "train_40317", "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": ["tooling", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a47dfb1109831e4c40d3aa436ef72fdc60b48026"} +{"id": "train_40318", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "documentation", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ef71baf0c9b2769f0de6c95fd1c0dacd3c551fb"} +{"id": "train_40319", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9492808c8825ff19f183cfffa2af785b37185ec"} +{"id": "train_40320", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b3bf32dd7de308b02d33916c00711fb2046644c"} +{"id": "train_40321", "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: 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", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9a6170ee81c0b09c4ae1d8b7439e59afaa0b53d"} +{"id": "train_40322", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "governance", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d01ce6445fcfaf5e6dbb4a99ac743f6f36621f67"} +{"id": "train_40323", "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: 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", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fd9701522ee87ed80c96bf84b1258d5abb6341c"} +{"id": "train_40324", "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: 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": ["tests_are_truth", "evaluation_metrics", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0112406420c68be32b4b953349d309a20e2b62e"} +{"id": "train_40325", "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": ["documentation", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbb0098d076b083686201037ed107f01b6e2475b"} +{"id": "train_40326", "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": ["reproducibility", "documentation", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7248a348e7c624e4dae08d2ac1bbde8f7379c27"} +{"id": "train_40327", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b04a0c1b0645a5d4222072492643561858f106cd"} +{"id": "train_40328", "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: 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", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f96aeda75995a6ed829778b3c9a89b2b0b7c9b4"} +{"id": "train_40329", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fc480daed913ddbe584c446fb9c6c6717420c37"} +{"id": "train_40330", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2f551e36627c08fab27d0ecc10a537d9b1e2adc"} +{"id": "train_40331", "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: 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", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58b7954a42fac6a2deaec3a2f577ae82c85d3c6a"} +{"id": "train_40332", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18be958d7eb730540c7ba063d8be1e64be049afd"} +{"id": "train_40333", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9106cedda961d2b9a28a2acc4fa43c96a3921fdd"} +{"id": "train_40334", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bba1d55dc69575c4d8bc9e378ae442bca071ce5"} +{"id": "train_40335", "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": ["documentation", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d20feb0cbd729e8de3ac332d3f13c7a82095f603"} +{"id": "train_40336", "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": ["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", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5abefee442e95ab905415390584114acf0b6a24b"} +{"id": "train_40337", "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: 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", "tooling", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bec76fc1c6dcc0d59aa884d417f23f6041de66f5"} +{"id": "train_40338", "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": ["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", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59531f312b0c45e45a1de03bb5da62513f495815"} +{"id": "train_40339", "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": ["documentation", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2531974312a81dbe7f6475d873732f4ff049349"} +{"id": "train_40340", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ec0686284865a507cc1785e8bccd7313dbbc65b"} +{"id": "train_40341", "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": ["reproducibility", "repo_scale_reasoning", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3717120ab2dc0e36f27149a46592688e4cb1f8ec"} +{"id": "train_40342", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "governance", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "611b4a0628c3fd3e1a2050c3d41d3e1758c00877"} +{"id": "train_40343", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5761dd242984e54b86ea846118964cd679b58313"} +{"id": "train_40344", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f410e55ff5320c50a37d4b6bed6ab3b5d1d70ab6"} +{"id": "train_40345", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "838847b678bb9a097cb4ad91188e85d54e4f730b"} +{"id": "train_40346", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a154f28ada9fac9773c1a5a06a1cbf91032b2fd5"} +{"id": "train_40347", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9baa4f3859943eb17b7137c8479a0a33e11af1ad"} +{"id": "train_40348", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cbcf45c9a48842a6d5c2bc1bfc48e02470dd8b9"} +{"id": "train_40349", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "639c20ecefec3ac934c0821b3420059f120109cc"} +{"id": "train_40350", "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: 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", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77e51cd749e5a34c2b980595e2ddf467dcf473f6"} +{"id": "train_40351", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f4bc176a47124910070f9d3450be12f0f20d966"} +{"id": "train_40352", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b072d6bc9b3f850b8dfb0c58f7b6e9afe4bad031"} +{"id": "train_40353", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd396ef70dfa8574fb8303ec99234484731a8f47"} +{"id": "train_40354", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4471b0d4dd0468312fb4699270824b1e983f6597"} +{"id": "train_40355", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd324f8b002b73ea46aae71063924c4f674827c7"} +{"id": "train_40356", "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: 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", "governance", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6aadce92c9cb280d76b208bcba576f801a681126"} +{"id": "train_40357", "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: 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": ["documentation", "repo_scale_reasoning", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37d88806045aca688c0cdbf064574db3aeed447e"} +{"id": "train_40358", "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: 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", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f10730ac326434c64e73d2b75d65f51cfd8ff3f"} +{"id": "train_40359", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08ddaee0f4ae2db5a7058ca2f3ae0f246e45404e"} +{"id": "train_40360", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93f7fec2edd316eb85fe7a24b98ad70679c263e9"} +{"id": "train_40361", "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: 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", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bae2c5198fa5d0a78a873024aab2ef5ddd852b08"} +{"id": "train_40362", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5015e61967ea98f0158ee4fe4a74cd43b851ccbb"} +{"id": "train_40363", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e79dac0a6946fb66030be12295e3fde044cea1a"} +{"id": "train_40364", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2c8f7135849ab2498a175f099e059f43e7f3131"} +{"id": "train_40365", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8fd50f2a0192b53ac21c963ee2de14541e270476"} +{"id": "train_40366", "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: 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": ["security_gates", "governance", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31bf7044a3baf09e6613d110d2027c5bababa45c"} +{"id": "train_40367", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "428e46238afc7a6eea0db6ff55cc08308635fc8f"} +{"id": "train_40368", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa72155f431139b68e2fa4d0408d21884ce85117"} +{"id": "train_40369", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb4e32c2f88f73ed91a6835b2148f55c2126eb7b"} +{"id": "train_40370", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ca2d53c78f894f7540b397677602cdf0dbb4d30"} +{"id": "train_40371", "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": ["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", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df73bdc1c252852e75ce13d0a88dc290902ffc75"} +{"id": "train_40372", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tooling", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2093c0b9d150f01ef11a898da3f13923f9105e1f"} +{"id": "train_40373", "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": ["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", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "896779aa8bc0725402b1ef746f8c4078313dce5f"} +{"id": "train_40374", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c41419c5eba8fa123e46e4e6524c739b7fc6a555"} +{"id": "train_40375", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f711b09e61504412f9c32ce18950264ab67b8cd3"} +{"id": "train_40376", "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: 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": ["evaluation_metrics", "cost_latency_tradeoffs", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04d8e5a7718048ffc89c46d01fd820c3bcace7cf"} +{"id": "train_40377", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0e4054eb016d294c90b9a5ee2a24332db80fdcc"} +{"id": "train_40378", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e3a825dfacb8f841feacbefb9263ab242e4ccff"} +{"id": "train_40379", "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: 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": ["repo_scale_reasoning", "security_gates", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93b7be84c69047d9f2a045db0db123ab3c1f59f9"} +{"id": "train_40380", "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: 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", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e70d0af7fbd5fdbf0466acaa4a5a4b03a579d178"} +{"id": "train_40381", "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", "security_gates", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcfdb8d7ab08340fec469705774f5564c9389778"} +{"id": "train_40382", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c978bcfb73f196447e70c38dcfc819ac17683a98"} +{"id": "train_40383", "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": ["auditability", "evaluation_metrics", "tooling", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66f61bcf44e0a6308b0b849d61fb1ec71b2e568e"} +{"id": "train_40384", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b430788b8062e49a7f4bf3a8ed01fb437d7fee6f"} +{"id": "train_40385", "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: 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", "tooling", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b44d2c6976236a6af62aee27aaee1dbd84716dfd"} +{"id": "train_40386", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "tests_are_truth", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64d47272dec5dbadd197ad3df2ea7084f50f1193"} +{"id": "train_40387", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e01f71883eb85bf621f33bb5c5a86fcc8669c7e"} +{"id": "train_40388", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7afd2007fda6c77ba634218eb304a1c087f02d5a"} +{"id": "train_40389", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "security_gates", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b87c1421a8e894a5bb715f21a56e7d8544b3478"} +{"id": "train_40390", "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: 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": ["security_gates", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f972ce0152b5990ad24dad02c9290ca41a804fc9"} +{"id": "train_40391", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["evaluation_metrics", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0ba3d34319b5a674e964f0fed48a697e98d200d"} +{"id": "train_40392", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["ci_integration", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4dfa6f26921698a392efd255f418b3146d1d2646"} +{"id": "train_40393", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6828a82fa981e31da9eea8549e70b9088d139d6"} +{"id": "train_40394", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9729e283a8930aa99b0ac5133ca6b78d6b0d7dd"} +{"id": "train_40395", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd0f56c60826f90c173b361c4c1449aed7ef2367"} +{"id": "train_40396", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec2906d5cbc80df24c049133658cb94517e0f92b"} +{"id": "train_40397", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["repo_scale_reasoning", "governance", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "807a46e4b6a7fb499a4926537c3615daee52ffc9"} +{"id": "train_40398", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f9583abec7a9b05f63041b628fd4341ff5d940f"} +{"id": "train_40399", "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": ["documentation", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ec30ef6bffb046f87b97f31ba838c47cc0d0f7c"} +{"id": "train_40400", "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": ["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", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "699663e1359c37be749705cbe447c9d45809f05d"} +{"id": "train_40401", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3eedd6a775eaa7a77ea03dfb8f4e1b9e79f7fb6"} +{"id": "train_40402", "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: 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", "repo_scale_reasoning", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86386c8e973d316a39ebc0fd3c9261b1773a0217"} +{"id": "train_40403", "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: 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": ["tooling", "auditability", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e537f7e6d39a5ca0cdc7b0dd014b01bcea4ba33"} +{"id": "train_40404", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "779abec667855149c4e5668621f5ea5629e43138"} +{"id": "train_40405", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2c498d4577e57e473355516f1fe254506412365"} +{"id": "train_40406", "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: 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": ["security_gates", "tooling", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a418cf60f5fd718ef6aa33e8aefb7c73c96d172d"} +{"id": "train_40407", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db4c388c50e1e91f700f5829c4021dbfbe7c2b79"} +{"id": "train_40408", "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": ["ci_integration", "tooling", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "699663e1359c37be749705cbe447c9d45809f05d"} +{"id": "train_40409", "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: 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": ["tests_are_truth", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0923183260a00f8e8323a216a2f386119955e7d1"} +{"id": "train_40410", "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: 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": ["repo_scale_reasoning", "evaluation_metrics", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1316c96bbdf28a7ad31d9ec8ac1f5e7165962c2"} +{"id": "train_40411", "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": ["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", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3259e656c0c3013c1944063645faa44979e26aa9"} +{"id": "train_40412", "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: 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", "tooling", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73794dcdff1eee73547bebac58d6e4720506b4ba"} +{"id": "train_40413", "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: 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": ["tooling", "documentation", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4017cc040b7b25939e496a707d706a3a49c5d824"} +{"id": "train_40414", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db643c123df2f94028cf593a78c4f79116d0f946"} +{"id": "train_40415", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec023554f72bc7a11533baaec74f01f76c721be7"} +{"id": "train_40416", "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": ["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", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e23d570cdcbdde1d5cf580fc74e7dd0fe68986a2"} +{"id": "train_40417", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f621e30bd75e0958c61a42d5a4b8627f8c1219be"} +{"id": "train_40418", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ac2084e33da878e6007536c59ca18596ef88467"} +{"id": "train_40419", "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: 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", "repo_scale_reasoning", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf87db71ac275f0a11264a686d137232e0129873"} +{"id": "train_40420", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22b56b0a98b30e86615e0810fc2dc7b2196edd9f"} +{"id": "train_40421", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cabfbea123bbc065711bc2cb527e2ed24028716"} +{"id": "train_40422", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddf466affafeffcb543738424d510d743b91ab7f"} +{"id": "train_40423", "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: 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", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e6d4d40c193f78d28bbbc8ba7bd521340390d62"} +{"id": "train_40424", "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: 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": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a8cb3d84606f87f6411dddec1eebed590982f6d"} +{"id": "train_40425", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81b97909062fbf3c9e068032042069a5d47c2d1b"} +{"id": "train_40426", "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: 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": ["security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "454fc72d206887ba8d3aa95e5345dc6d6385c5c9"} +{"id": "train_40427", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04401ca1cd0baa2f80afffb89fedb824d9fbd529"} +{"id": "train_40428", "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": ["auditability", "tooling", "documentation", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "213bb6bfc016a1436411f830fa772820bf1b4cd2"} +{"id": "train_40429", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61647b78e291e2fb0fa9552d35d0d274ef3a0ac3"} +{"id": "train_40430", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45497913664d77503e321794f003a260b6524a76"} +{"id": "train_40431", "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": ["tests_are_truth", "tooling", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "969b47d187e87cf0e2e1786d158ec4034523d8b2"} +{"id": "train_40432", "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: 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", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7c3560c704ac859647a6b6f843d5397b37014cf"} +{"id": "train_40433", "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": ["reproducibility", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87e7642838774dfd635cb91ebdb3fabd23c550a7"} +{"id": "train_40434", "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": ["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", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22f42f8c834e48a87b24b981d625f0b5e1f82e98"} +{"id": "train_40435", "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: 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", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19a21bf75370ea6ab04adb6b2ab2b6681eeadce8"} +{"id": "train_40436", "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": ["documentation", "tooling", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cca13bcf6835e2cbf3d86a32388453c454f7dd16"} +{"id": "train_40437", "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: 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", "ci_integration", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "045b4a07f3131d39038e86fee32a79a2067b8709"} +{"id": "train_40438", "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: 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", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd7cc0fe1a3c79dfdbd0833e36d8f94dcb27f829"} +{"id": "train_40439", "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": ["reproducibility", "ci_integration", "security_gates", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d6304a6f536583ab57b342a59cf5a2f2feac519"} +{"id": "train_40440", "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: 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": ["tooling", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "851491f0b7253cc9457031cb756427de7fd88c12"} +{"id": "train_40441", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2380d205fbac16c0c65f56a72e878727929cf4bc"} +{"id": "train_40442", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57741dcf1d88d5fca84d199d803f90635d82ff8f"} +{"id": "train_40443", "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": ["governance", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f84af4a6996eeb4976689b9186fc2407c97e196f"} +{"id": "train_40444", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["tooling", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b510919f7f3cd3b8759d5d49ddd39e15c94dcb3"} +{"id": "train_40445", "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: 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", "cost_latency_tradeoffs", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1aba38646dbfafad9ffca0616a7aa873daf2a291"} +{"id": "train_40446", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16664514e23be040f46e7d846338ebd262ee20f5"} +{"id": "train_40447", "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": ["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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b22b4b49fe844f469740b8a414868b0e395b56d"} +{"id": "train_40448", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1d1e10ef17e57ae88f7297b3cbcbf8cac7d35bb"} +{"id": "train_40449", "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": ["tooling", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f6b99c1e81fb65940d76d693d4be0ce9bf42836"} +{"id": "train_40450", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7b7698c48f20d00593097bb7b075fddb0cb8256"} +{"id": "train_40451", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "governance", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eae9cd0d1801f5867fda163511603f1287338d46"} +{"id": "train_40452", "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: 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", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "219e41510e2b5bbb1d803128816b4afc6c88f828"} +{"id": "train_40453", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "reproducibility", "governance", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5e5d645d2b73be8497b84144fe95fde4baa5ad6"} +{"id": "train_40454", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5ecbb81a5f9a9856fb0b2b49a5f539631274359"} +{"id": "train_40455", "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: 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", "tests_are_truth", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0275b71c53e24ad8425d520b4406cddec9a7ec9"} +{"id": "train_40456", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a906d93ac68ffb8cea18cbfa64070f398d78c2ff"} +{"id": "train_40457", "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: 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", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3dcd26293146240bfa733344753e1468d3da1585"} +{"id": "train_40458", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b03199aed429c795494b32a196c9bfc957a9075"} +{"id": "train_40459", "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": ["reproducibility", "governance", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a418b363c637ed3e16c7678f2dad36ed537d74a9"} +{"id": "train_40460", "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": ["auditability", "repo_scale_reasoning", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1bc4eb6c3862f829116c7c63e568fad0f7a7f113"} +{"id": "train_40461", "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: 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": ["cost_latency_tradeoffs", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc6802aa050f3c2df011b995af0b277535b79fda"} +{"id": "train_40462", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64ea63031cd029b31a9188ad28afdde70bc276fa"} +{"id": "train_40463", "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: 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", "evaluation_metrics", "auditability", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56f52de8eb80d684805e7a273ca0d45b0483f1b0"} +{"id": "train_40464", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08fa1e68144fd0f67a7cc3357753765b0f922097"} +{"id": "train_40465", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdf523c8fa69c36bdf9f2b6e44ae0bb96fec4605"} +{"id": "train_40466", "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": ["auditability", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64514e82b2225fbc8ddebc8a738a55e84a0b2664"} +{"id": "train_40467", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa4593d51bd89e7502ee192335859267199f239c"} +{"id": "train_40468", "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": ["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", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b0c32c134fb602bd452ee79258c1aa3b5d2b2d4"} +{"id": "train_40469", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d87b4d3c67d7411cd6c6e7af0307b7ce09164859"} +{"id": "train_40470", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7dc453840417281d6bf5290833b769021f785ac"} +{"id": "train_40471", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready 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", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1be08020ea60e1bf21bc98044e89e1ad2d8f7939"} +{"id": "train_40472", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "695ac3cfa5c8d7b2aaf6ad107183be0461fa3457"} +{"id": "train_40473", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5cd7c21a06c594ace42833d70c7d273ddd049fbf"} +{"id": "train_40474", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1753ad5d5602be652ee03386292828a85c7b19cd"} +{"id": "train_40475", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4303ff25e9cee55082fceafe936a515377cefc0"} +{"id": "train_40476", "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: 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", "security_gates", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58497f74aee9ae46e96c6214ec1b69ac0d4cd2fd"} +{"id": "train_40477", "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": ["tooling", "security_gates", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27b20287809be0d9d5f5074a855d02ef91b91215"} +{"id": "train_40478", "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": ["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": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9eba037f5528472a0bf10537dd2a71041cb1c204"} +{"id": "train_40479", "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": ["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", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc60411947b1685d6fb6139f600f51860a4dc263"} +{"id": "train_40480", "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: 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": ["documentation", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "936781168d13d124ddb0f30ee1f31198fb52f3c4"} +{"id": "train_40481", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f9583abec7a9b05f63041b628fd4341ff5d940f"} +{"id": "train_40482", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["auditability", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b30728a03c77a55ee2c129d143beac08f9edcde"} +{"id": "train_40483", "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", "ci_integration", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "243893d38ffcee49347a8e012b134f3e63b388b5"} +{"id": "train_40484", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8cc6efb179dc3013bcff1be76a74921543ef5f5"} +{"id": "train_40485", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc5054e4612c42f0d4762273ce20477edce6a0a5"} +{"id": "train_40486", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "712f1790f1e386981c88d1c32d94dde20b51fb8b"} +{"id": "train_40487", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3f0170024180df9b183737f9e299fcb93476db9"} +{"id": "train_40488", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "008c75d926378ea9c5772fe6b62dd613b3b39c06"} +{"id": "train_40489", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "reproducibility", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd092b753e1edf209bf24f0828d317ff522d7cd5"} +{"id": "train_40490", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e304653be1dcad4c4da8d4ae5173184b8338b4c"} +{"id": "train_40491", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb77d30555b74e0a83c3d7f1919b2f08979d052e"} +{"id": "train_40492", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4348e2ace761aaa5df8acf931d24fd2639d928b"} +{"id": "train_40493", "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": ["security_gates", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "511bb8241930aeb214d399bd67edceb5e2330d15"} +{"id": "train_40494", "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": ["cost_latency_tradeoffs", "security_gates", "tooling", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b60acde04ac653389b01402d11cc5ea8f6531c7b"} +{"id": "train_40495", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15e220933f51ecc8887134b95e1fcb49d09e1dd4"} +{"id": "train_40496", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5970490cbaf4b0122c005fb02294771a7a074ec"} +{"id": "train_40497", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "155b241bd40e1c013e5f77fad13019059b2640b2"} +{"id": "train_40498", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f46441a94792cb47d69fc6c44a760559da5a552d"} +{"id": "train_40499", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba778dcd4ba256797c847a75355cb3439c8e046b"} +{"id": "train_40500", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1aa8fe4c1f39f8a7a0827e65b5a3a501d975e896"} +{"id": "train_40501", "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": ["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", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f93cd7f2c3ad71cb5c7c2a79d90352f21ada4eba"} +{"id": "train_40502", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e34003a28d0f2eab6d7875e1e142b04849bbfee6"} +{"id": "train_40503", "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: 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", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14992629785714094062c32a5a15c0b96c8b2d2a"} +{"id": "train_40504", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b84a6c1e8b547934f673378ad12974c7100bb3e2"} +{"id": "train_40505", "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: 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": ["security_gates", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff4ae34cfc8463732f4fd97b6130a3ce537f676e"} +{"id": "train_40506", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac18068b90b292481b365fdcdd49c247dd2fbb47"} +{"id": "train_40507", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a390e646bc947ffd7d609ed3525f01581e6394b"} +{"id": "train_40508", "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: 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", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a0a02ee3ea09a77ac45d9e368da9e9f258c5aa6"} +{"id": "train_40509", "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": ["tooling", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55e7003b115a0acf58dec68f75d1d67a929894e9"} +{"id": "train_40510", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9e84b96faf27173e81cead4ba9ce009159f6620"} +{"id": "train_40511", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "cost_latency_tradeoffs", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f11136efff252704374d1467638a3530f52fffb1"} +{"id": "train_40512", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f0b0ace0b7e4be9014f98c1645d5c7482194589"} +{"id": "train_40513", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b7ac913b5f1ccd2e32a7cfcbf7b8999c7f91c40"} +{"id": "train_40514", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3630c6c697a256a280dca12fb0cae665339414f5"} +{"id": "train_40515", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "auditability", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "042e7a645ab11133703c8d67a7b73f3ba38835e0"} +{"id": "train_40516", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "932312e472b734409c0036b0abc3e7894e991610"} +{"id": "train_40517", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1b97b698d8e7274fa49a4a0d5a966c1590965bb"} +{"id": "train_40518", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "231e878e8e14bd81834f5a2a65abf969d2b268dd"} +{"id": "train_40519", "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": ["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", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ede2f21e135c847c3718a1120440138d7b9c9fd"} +{"id": "train_40520", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "auditability", "security_gates", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cd87228170de57c5f25893b4f7cafedebcd6514"} +{"id": "train_40521", "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: 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": ["auditability", "evaluation_metrics", "documentation", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "331b209a41b72f07d40a6bf5bb5abd943ae59b1d"} +{"id": "train_40522", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "tooling", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b2f870f331ce446b90a036f8e385aaca321c261"} +{"id": "train_40523", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e3a825dfacb8f841feacbefb9263ab242e4ccff"} +{"id": "train_40524", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afd47a83065b2de86807ec37f9a1db1d6fecdc31"} +{"id": "train_40525", "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: 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", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01962f28c95ede619e9616f454d986e8a3ee5c14"} +{"id": "train_40526", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "documentation", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "622e73f5b6ea441d46f69fc0cb4de8433f06f817"} +{"id": "train_40527", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["tests_are_truth", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10da8a8e5153ac4cf8949d9c6762c8120854fa7d"} +{"id": "train_40528", "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: 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": ["tooling", "tests_are_truth", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcd4534bfd0227f4f1f30ef370ef6d4798a00c52"} +{"id": "train_40529", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "570fe076ccbcf7fd3361adc52a612633bcf0745e"} +{"id": "train_40530", "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: 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", "security_gates", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c43ca0779f00e5e6be35c428ed0df1146f13ee76"} +{"id": "train_40531", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92e90975220349cd88504c4fc7f30efb76f3f1d4"} +{"id": "train_40532", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dafda0f5e7844ac27f64e732cf97308ea99e51d5"} +{"id": "train_40533", "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": ["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", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a68f948b9301ace61336c153a9ad16be9e33ba6d"} +{"id": "train_40534", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fdf779c2d30d4896a2755cc56313714fc637961"} +{"id": "train_40535", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "security_gates", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21849a7c820160f4ba460b6689e707683e6553f0"} +{"id": "train_40536", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0bba5703e43c7dba01893587c84a2d651e392f3"} +{"id": "train_40537", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a79a529222be2069abe578589db2a69ee3e11bb"} +{"id": "train_40538", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a4ebd2f2b2cf3c09c61e1cee6d53a50545be76b"} +{"id": "train_40539", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "documentation", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37646f521784ab5690d008807359bcbcfdede18c"} +{"id": "train_40540", "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: 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": ["tooling", "reproducibility", "auditability", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60958d880dc1996b29cc6aa2a05a4bd9dacd0027"} +{"id": "train_40541", "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": ["governance", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdf2d661b99fee070240ddcfb81d8609a1fae1e9"} +{"id": "train_40542", "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", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "147b122748e928d7ec24c864c5225673778d6b58"} +{"id": "train_40543", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36cf90eb0ef21c54ce52a42ad2ff573b1fb0aa1e"} +{"id": "train_40544", "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: 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": ["tests_are_truth", "governance", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbde2cb63d83092ee5faeacec468d08d3732e2e8"} +{"id": "train_40545", "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: 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", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44cd33a06426f971bfaf580c57a932ce7b564a7f"} +{"id": "train_40546", "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: 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", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6d2506a7c024aec78b3a3172bf8de5c98bb98ce"} +{"id": "train_40547", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c99db5669b1592b572f35ee83591b1d31e1313e"} +{"id": "train_40548", "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: 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": ["repo_scale_reasoning", "auditability", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3c4d68ba32dfbabd4d245d8ca5e6c2489730d15"} +{"id": "train_40549", "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: 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", "auditability", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0698fa67edd5751fbf83251bd1b07aae790c155a"} +{"id": "train_40550", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25b3752fe34f428df7c1010316a48930947a6157"} +{"id": "train_40551", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2d1334b1b79c9ed0cff9a22a86d49679ef041b9"} +{"id": "train_40552", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "832e81862bf341a4e15d76c13d7b78e74e3780ba"} +{"id": "train_40553", "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: 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": ["repo_scale_reasoning", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e54526e7ca40dcc6d42a83f338e8adece31c8ec"} +{"id": "train_40554", "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: 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", "auditability", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83e74998f54d30073e2cd7fd90ddd17e93eb396c"} +{"id": "train_40555", "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": ["governance", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af47f786a2f62667f44778f51b47dcba61559ae0"} +{"id": "train_40556", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "security_gates", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d9624ceaac84d4a9a40c2b791d4d7589f2a6b94"} +{"id": "train_40557", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f04640601407720d4ad887aabc96bdba067bb1bf"} +{"id": "train_40558", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dee8dd69b18ef45a1020b5de1902051aabd5be5c"} +{"id": "train_40559", "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": ["auditability", "repo_scale_reasoning", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3a7be1615f13557524072c9df6ac0eac07b9b12"} +{"id": "train_40560", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ff9c40b027fba141b02f92bcdb88ca88d9247a8"} +{"id": "train_40561", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f55d13cbc904691c51ece1bc78d41114c936545"} +{"id": "train_40562", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "security_gates", "reproducibility", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98cc578131a7971f85edc852e3d6d0d3c9937da4"} +{"id": "train_40563", "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: 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": ["tests_are_truth", "evaluation_metrics", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee25788ef2bb6602a7ab001b2b15a3d10fda78f0"} +{"id": "train_40564", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37677db34224740be552276e556b95c6f39a4972"} +{"id": "train_40565", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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": ["security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76becc765b4c2d3b08d3d4f8bb51364a90e82ee4"} +{"id": "train_40566", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7543a8ae8973a57de726d90ae36f763a5797de8c"} +{"id": "train_40567", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2372b8b849a8ed98be40c4898503da8a418e1a9e"} +{"id": "train_40568", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4348e2ace761aaa5df8acf931d24fd2639d928b"} +{"id": "train_40569", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49cfe255142361260891ad0fd9e20dbdec822768"} +{"id": "train_40570", "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": ["auditability", "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b79531042cf95a55db5ad90bd49d8a642aed2686"} +{"id": "train_40571", "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": ["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", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "faf3a099522c3c907843c5c95dbaaa0e33b99f75"} +{"id": "train_40572", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "security_gates", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0746a0b81e7f1882159cb5d7f3ef49a6937ccad2"} +{"id": "train_40573", "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: 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", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5caa926cc2ad543ac1693d69e59cda05a28dd83d"} +{"id": "train_40574", "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: 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", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e97de48362433e4694e263c3872218bc4f4d76e"} +{"id": "train_40575", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5d64e0e0485f93b1de3882c078ba8c90996ba0b"} +{"id": "train_40576", "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: 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", "tooling", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "008ddd2d2762d6477f159a2a97d21b571ac32c8f"} +{"id": "train_40577", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "documentation", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76a49a9c37ce23d193c650d67abf73afdd78be2e"} +{"id": "train_40578", "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: 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": ["tests_are_truth", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b742721083fad8e02cbbef4e5eb0b561afa09d15"} +{"id": "train_40579", "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: 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": ["reproducibility", "governance", "security_gates", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a31eeecf189bbeb9c5e0f048f7ffd49ff25a7208"} +{"id": "train_40580", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45b5a58d27d543a3d77bff4184afad49f09f5f66"} +{"id": "train_40581", "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: 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": ["tests_are_truth", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "638dfec90067fb5c8bb565b1a354bf911cb709f0"} +{"id": "train_40582", "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": ["documentation", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0fb01a19fb7295182706fcce6e88f7c4c56d2aa"} +{"id": "train_40583", "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: 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": ["cost_latency_tradeoffs", "tooling", "reproducibility", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c45fbb2b432e148f0f94c3cf93c4bcd587ad86c0"} +{"id": "train_40584", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa1930c19cca9f0e44f15b2e2aec120af9bb38c4"} +{"id": "train_40585", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e96c292e8e4870b3f1196c085c8d06b0ddc2535"} +{"id": "train_40586", "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: 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": ["tests_are_truth", "governance", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee56b77cbc50434f13487ddaf007131bdfba0beb"} +{"id": "train_40587", "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": ["evaluation_metrics", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e6c9ade6773c25459568f83122a232d5c211741"} +{"id": "train_40588", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "auditability", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a611c4e77331e03c19af4f5f7dda191a2e862448"} +{"id": "train_40589", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae556a09e86d9170118acc22dd349363428fc783"} +{"id": "train_40590", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3944e6f1afcf5792e6ac3b269b32cd5927420d30"} +{"id": "train_40591", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6682f2687371a6e8b2d1570e1c5c8d2dcc6bbb2"} +{"id": "train_40592", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["ci_integration", "tooling", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1a9822314e6ffd3b46e6c659fd36f21f3c94de9"} +{"id": "train_40593", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7746af5b39c606c7bf36bfb082800d0d4085911"} +{"id": "train_40594", "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": ["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", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b360361ef83ab5ddefdbefa10b8e83892d0f42a5"} +{"id": "train_40595", "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: 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", "tooling", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a455a6c64c7a2de4cac27e3b3c6781a6bd96a8a"} +{"id": "train_40596", "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: 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": ["auditability", "security_gates", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d51c7c9607d4161608f0b30bef6033cc550bda59"} +{"id": "train_40597", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["governance", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12feb66bcb93ec4e1cb81a5fa985395e2952355d"} +{"id": "train_40598", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "security_gates", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8346334251620af16454d8a44835ed31d2d31b52"} +{"id": "train_40599", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba92fb3e5c4d76def32cc8e57c6dce4e9c28294e"} +{"id": "train_40600", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20180cae3acf460950680692d2d754542532c181"} +{"id": "train_40601", "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: 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": ["repo_scale_reasoning", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7ef2ddbd4b5d7974c20a87b5c417c975933c581"} +{"id": "train_40602", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4b98d32a3a8e8ce589d56167cb247a9e846afe6"} +{"id": "train_40603", "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": ["cost_latency_tradeoffs", "tests_are_truth", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c63c9102fa395af58cac038b924c0e5fa93cc6de"} +{"id": "train_40604", "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": ["tooling", "documentation", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b9b4e61e80f10413c88d26592b9fbfe50dc9092"} +{"id": "train_40605", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8346334251620af16454d8a44835ed31d2d31b52"} +{"id": "train_40606", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a3139fd16430586cfd96540efd903274520e562"} +{"id": "train_40607", "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: 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": ["repo_scale_reasoning", "governance", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9472e173e180705b19d2d6b9f25fdfcc4d821b0e"} +{"id": "train_40608", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "documentation", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9ef5aae2fd80ce5805041e775d1d6e58fe0e096"} +{"id": "train_40609", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "governance", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ebb4479628650fafa6d6580d390aafb78d47d05"} +{"id": "train_40610", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5d59b331cde9caa59d7c47315090ff8aeec6c6e"} +{"id": "train_40611", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "907baf22d69990a88f6bb82094d6f017cb71e8f6"} +{"id": "train_40612", "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": ["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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0339c84f5835afa3decf1f286d5387c1d1ec7779"} +{"id": "train_40613", "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: 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7804df1da499a7194d8abc14c6534e13452e4415"} +{"id": "train_40614", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e625b133e8b2f6cee1844db85794ed9af00cab81"} +{"id": "train_40615", "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: 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": ["tooling", "reproducibility", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6663742cb7a812d9bb868ec81193c11ca905495a"} +{"id": "train_40616", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "auditability", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fbf83c4e6e0bc2b06cca3e8cae52ad4d3a4ac2a"} +{"id": "train_40617", "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: 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": ["governance", "cost_latency_tradeoffs", "auditability", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da982dd89d00c925d967c326b89c969622cd4ab5"} +{"id": "train_40618", "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", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a7527abe57685af57cf511e8f2803074a9c19a2"} +{"id": "train_40619", "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": ["tests_are_truth", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fea0bd741622a3dff2cb38856dc1b49232f37001"} +{"id": "train_40620", "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: 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", "tests_are_truth", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de1f0da187bc60c90d7d4322365effeb8744766c"} +{"id": "train_40621", "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": ["ci_integration", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc682a53d95dc32aee5ce70b017aad3fe88a7203"} +{"id": "train_40622", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2dc3833ba7f87e8fd8db50be4907316aee5bd682"} +{"id": "train_40623", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f5326c774ff3f2c6f6f8e0c0f3ddb0b206f6400"} +{"id": "train_40624", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["evaluation_metrics", "governance", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d7e995764e375ec45d043e84b8ab3ebfaa11649"} +{"id": "train_40625", "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", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0125a48d7a92afb403a7d162ea2df7df73c29f9c"} +{"id": "train_40626", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64c92f6e991047881e16ab70f7261b22c9c9367c"} +{"id": "train_40627", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "reproducibility", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9b47b9af42f23688097c83c8cee5831480e3d55"} +{"id": "train_40628", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9ce05fda1f0bd8d87b1ad59016d583307124232"} +{"id": "train_40629", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e2e837f64dc8163c4ca44d0e3c28ad4ab6ceb95"} +{"id": "train_40630", "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": ["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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39098818073d0ec670c9c9053b48ae13697766c3"} +{"id": "train_40631", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29f24e3915e1a66d7bd497d4db10da6d1de38258"} +{"id": "train_40632", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58d690580989a3acb127bf3752871f7cbf21c8a7"} +{"id": "train_40633", "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: 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": ["auditability", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0a27b9298748f87f9b797a9dfb4a2491b5cd80f"} +{"id": "train_40634", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92c16a6df18c7bdcdf9531d1433ee59b624344ff"} +{"id": "train_40635", "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": ["governance", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78249f821dcffa92a3d1ff9138a5398a5367f08c"} +{"id": "train_40636", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd3553e1be008167476844e3deb95d2afd85bfac"} +{"id": "train_40637", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "auditability", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6936664fa8483934989ae61f9efe413a188d138"} +{"id": "train_40638", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "003dd16a3f8419bba77091e8280f1d7ae5951762"} +{"id": "train_40639", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a1b4048e860b27d5279cc7da6a6298c34e70de9"} +{"id": "train_40640", "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: 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", "security_gates", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d7c815c04c0fe9b3599122b58cbe545a09f62b7"} +{"id": "train_40641", "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": ["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", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7ed5cf9c0c084f399d7eafc9749d5de38ce321c"} +{"id": "train_40642", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34d726b2648eb29dea639058c967fbc4295d76a2"} +{"id": "train_40643", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f75193d1bd21d4419b6995d38305967e60240c5"} +{"id": "train_40644", "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": ["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", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ba27224adda7f3dcf70b6be91be72ca950a1b41"} +{"id": "train_40645", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb67a2121ec4ba5e1838fdb95df761d8f41e0756"} +{"id": "train_40646", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c19959216959b58b6036b7e33ad0f906d15a751c"} +{"id": "train_40647", "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: 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": ["ci_integration", "repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e6f602ce5c6188f8ec037eb8327438c827741e7"} +{"id": "train_40648", "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: 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": ["evaluation_metrics", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30a63cfd52e893a05a4b59ed924a5d5ce10f5218"} +{"id": "train_40649", "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: 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": ["auditability", "security_gates", "reproducibility", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48e14be1c2135d0145eee3ebacf2b713a337e4e9"} +{"id": "train_40650", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d674d3723c9cec1cc800efdef8cded169ed3954e"} +{"id": "train_40651", "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: 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", "ci_integration", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1348e4619358cc908e724c23b45487e75f80faa2"} +{"id": "train_40652", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1acf6b40ecf40315cba9f1402118ad4f26388b76"} +{"id": "train_40653", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "auditability", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61d73761fc5f3ad47c90eb31a40a8c95cea0bc25"} +{"id": "train_40654", "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": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43907aeabd79a6856396c895e701606b70d61638"} +{"id": "train_40655", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "tests_are_truth", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5007e4d8d36b22e40260f5e55ac8d2d685514d6c"} +{"id": "train_40656", "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": ["ci_integration", "documentation", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8f816baee979bac83953ecd8d90f1f18c500a9b"} +{"id": "train_40657", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ced6b4e85f266dc5ce408990470b5932a17377b"} +{"id": "train_40658", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4b3257b50cfcdf01d31797d0748262ca5a22913"} +{"id": "train_40659", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad9abd7f58c33c3dd3db8fb10a1d0e8b04db01c8"} +{"id": "train_40660", "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", "governance", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fceae268ce611d1920122539cd25805ad8f2720"} +{"id": "train_40661", "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: 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", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "794f6a7b95adba75e9e83ef2e68a7ef24c2ba66b"} +{"id": "train_40662", "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": ["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", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c21565a76dea538e7383e54bbd486a0a35f5b9b"} +{"id": "train_40663", "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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c0cdcbac8d0e20fc6348af696e674096cc692cd"} +{"id": "train_40664", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9b2e23698f62bdf4cfdc18364dce65c8b4f2574"} +{"id": "train_40665", "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: 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": ["cost_latency_tradeoffs", "auditability", "security_gates", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a61c7f734a478d547cbd9a72e91341f411d9570b"} +{"id": "train_40666", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66fb984dc946883c5d3659b360d907071bd298a9"} +{"id": "train_40667", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08fa1e68144fd0f67a7cc3357753765b0f922097"} +{"id": "train_40668", "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: 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", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "deca62b04555731a07efb421035237395e0b9cfc"} +{"id": "train_40669", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa82d614dea88d39983e08bfecf000d5b1386639"} +{"id": "train_40670", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c08b39d1de3e3d0d0729d466acd2f12d6e1b39e1"} +{"id": "train_40671", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb1cf62ce85b77364d632fcef9fe9a3712c674a5"} +{"id": "train_40672", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2444eb9074042830f72c8e15bafc69abedb68106"} +{"id": "train_40673", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4912312a2dce020b4c1d6871f64edb53719f63a"} +{"id": "train_40674", "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: 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", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33ae5fb881442bf95cc381258153e2379128014d"} +{"id": "train_40675", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "350c1d7e92ee380a5cace7ff37d760e152209423"} +{"id": "train_40676", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "security_gates", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "690b0a4b256afeaa8a702384a50eb075e7d0292a"} +{"id": "train_40677", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9eb2929db47bbe10668646a935a67afd3db4934c"} +{"id": "train_40678", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98ac3c02944bf7fa9aeba486f3e991676c0201e3"} +{"id": "train_40679", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af4f2981259e3ce5d12e28ccb4e6b9fef65c0801"} +{"id": "train_40680", "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: 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", "tests_are_truth", "ci_integration", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eadf88c1732253ba17f341a824dd9ac7509b01fd"} +{"id": "train_40681", "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: 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", "governance", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25f018204edcf1b0c4c0569f6bc40ba7d0f15630"} +{"id": "train_40682", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4bf5b91d61ec698ae25cf0ba10bf926f36e24a1"} +{"id": "train_40683", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "security_gates", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c91d6172e1d30d408e3e814dbef7e1634f2bf7e3"} +{"id": "train_40684", "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: 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", "security_gates", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "549db2b2d1c5ea6372ab52672d077a048ed9303a"} +{"id": "train_40685", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49a0140da7f00da31d3fd34bc12389dd311bcc6f"} +{"id": "train_40686", "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": ["governance", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d9dc9ee418c302de18fce55eb33852975657397"} +{"id": "train_40687", "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: 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", "reproducibility", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5711bc29ac6c15ed4f45dfc8da30cdd7c431aea5"} +{"id": "train_40688", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10222038612f554beed070308dc42892439a5dec"} +{"id": "train_40689", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbbd2be0c5149beff84b8ed2140a18bf95b19b4e"} +{"id": "train_40690", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74af8307111b9be833afd74597e33da0f1df0feb"} +{"id": "train_40691", "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: 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": ["governance", "cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "848e51e88ccfa2f1e1bf93e420278c03d59177eb"} +{"id": "train_40692", "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: 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": ["ci_integration", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7c3560c704ac859647a6b6f843d5397b37014cf"} +{"id": "train_40693", "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: 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": ["auditability", "security_gates", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68916c151b27f29a0ff71bdcbf0afa8a62739a4d"} +{"id": "train_40694", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18c3a39c067612dd1771629f70767e16e3949f47"} +{"id": "train_40695", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16f33fb57db0d4f2809d7d3a9b214c13856d405a"} +{"id": "train_40696", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "741b67a9cf16f70bfd54db1626626cd92f6b3787"} +{"id": "train_40697", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d36d3cac495c0399cef57c42c72a5a790b24f678"} +{"id": "train_40698", "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: 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": ["evaluation_metrics", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e3c23116b9d721683ea02bb7d4e7cdcf860728d"} +{"id": "train_40699", "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: 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", "auditability", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71acb3712983e6a69c5a84393630376261e7eabf"} +{"id": "train_40700", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "971a5bbbcdbc631a5a542818d3b157b5971b8407"} +{"id": "train_40701", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51a11675b4682715d495495f667c06d81c945e1c"} +{"id": "train_40702", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "documentation", "reproducibility", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b38df81d6e323c7b852da2a969ea859e5cdfb8c"} +{"id": "train_40703", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fb178cbb11610ba9f95f5e9169ad1188ee29858"} +{"id": "train_40704", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be15e2cc4b72ffc14ab16d872a2fdd1475796067"} +{"id": "train_40705", "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: 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": ["ci_integration", "documentation", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "417005f437c0c6dad4f359e90b637de1fed9e44c"} +{"id": "train_40706", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "abfdf3622bec7caa80a036d6dfcf0122171bf024"} +{"id": "train_40707", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f88163b1b1896ae680c25d6b88d185697d6e045"} +{"id": "train_40708", "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: 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": ["security_gates", "reproducibility", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bbd9757c3becda474812f237c5623d8fd742a7f"} +{"id": "train_40709", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0821df03b2ffe578d8976aa747b277dc1e48c88a"} +{"id": "train_40710", "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: 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": ["reproducibility", "repo_scale_reasoning", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "758412a33208cc61340f085db01f21e30c9cef05"} +{"id": "train_40711", "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: 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": ["tests_are_truth", "auditability", "governance", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1608f766c67427819d0f270573b34d8d54423ebd"} +{"id": "train_40712", "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: 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": ["cost_latency_tradeoffs", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fad3e4e0fc3b78ed6256f537d04e533187792229"} +{"id": "train_40713", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e3a825dfacb8f841feacbefb9263ab242e4ccff"} +{"id": "train_40714", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69152bb616fd177ff3439382390caf1f495b2840"} +{"id": "train_40715", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "evaluation_metrics", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b3b39d29d50323687f31cdd02ba47dc558580b9"} +{"id": "train_40716", "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: 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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f4d355f678d19bacd2909a640d13c5e054db8e7"} +{"id": "train_40717", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dea5727ad006605ec997bc53ca54bc0aaf76dfa"} +{"id": "train_40718", "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: 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", "tooling", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebb66e87afbb0500a942e54f05daca780f05dc83"} +{"id": "train_40719", "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: 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", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c462afa9b9fb2e964ca3e4126ca73c5f5100bbc"} +{"id": "train_40720", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e17ee16e39a02049524db9a6fafda50b1c43d02"} +{"id": "train_40721", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "governance", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f2a9d4574e558bee09e36b201b58563723c9cbf"} +{"id": "train_40722", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ce8fb1ea4fc27488c9951d3a1ef19876129ab81"} +{"id": "train_40723", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dbee4093f28c1f2eb8e9b89efc886a5c053a78b"} +{"id": "train_40724", "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: 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", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d91e7971dc03d84142e1e6abdbca7e8a9a1ebbbf"} +{"id": "train_40725", "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: 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": ["reproducibility", "documentation", "governance", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b59d1bc448e936582d93e2a782a68cdbc4e3068"} +{"id": "train_40726", "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: 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": ["documentation", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82165521e798a6c74ba1c6bba49acd35b5b95b2f"} +{"id": "train_40727", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd446a81c46c72d65cae15c330e7466654face24"} +{"id": "train_40728", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d55d6d8001ebf4ff511a7786c815bb7bedbc0ba1"} +{"id": "train_40729", "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": ["ci_integration", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ebeee507d1099954739d0f31b7861c768cbf232"} +{"id": "train_40730", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b3bf32dd7de308b02d33916c00711fb2046644c"} +{"id": "train_40731", "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: 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": ["tests_are_truth", "governance", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5031cca6d939ea56c1fb0b958cf1083a2c477fc8"} +{"id": "train_40732", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "ci_integration", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88dd3284723e036e0fa4ee9b4211be1a70600250"} +{"id": "train_40733", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d651ff4c0dc0eaf80178e370afb40fd4ef6dce14"} +{"id": "train_40734", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43ae297fdf9cdc060ae9798439ae1c12dcccbeb5"} +{"id": "train_40735", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6775dafb5136ebf7b3a1babbbe894954c9ea5fd9"} +{"id": "train_40736", "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: 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", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9357ab97ec96cee1e0801bf50cf85d26d2f9a210"} +{"id": "train_40737", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c3bbf1e5e9b5ed71de8b1ca461b14165e054787"} +{"id": "train_40738", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cedd3f02677957612d33a08b625642e6a06a344"} +{"id": "train_40739", "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: 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": ["evaluation_metrics", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a713524317b97115197d8f1e09946ed95969bfc"} +{"id": "train_40740", "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: 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": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1058f5215024e94943ffea4f91d51f1af4b1233"} +{"id": "train_40741", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "ci_integration", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5326b22a4a0d16f0d396df9c1d4247fae5304712"} +{"id": "train_40742", "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: 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", "ci_integration", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "364c5f8ff755b0fb792637d536a27f2ccf99fbda"} +{"id": "train_40743", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "959e25cbc6d3d71cd6d133ce456298ec75cd482e"} +{"id": "train_40744", "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: 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", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3be3770d2d0ae3cd7b3aa689e1d25db7830df24"} +{"id": "train_40745", "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": ["tooling", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f82276d99ae592fb16addd9df7c88f70ef1db2ae"} +{"id": "train_40746", "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: 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": ["documentation", "tooling", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74286fd36a3aa32e97351226f5c40afe4c0709c5"} +{"id": "train_40747", "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: 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": ["reproducibility", "repo_scale_reasoning", "security_gates", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0480ad0a4e03c294bf2cf23f49b2b922a8a77806"} +{"id": "train_40748", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8a6a9cc36ec6cf8ffc69156dc3ee2a7f6052154"} +{"id": "train_40749", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f93c199c6c00f32b6c1b4823079d93dd187b4f7"} +{"id": "train_40750", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f12833b72ac78cef2ba48f7e285b4f2792f7ec3"} +{"id": "train_40751", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1feb7eca9b21012c1c2f1e7f1f83a346a75b5535"} +{"id": "train_40752", "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: 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0ab6e6b1a075de440c9f16379a88f9bf5a9c409"} +{"id": "train_40753", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eadeb0945349df0f6d57f71125011f339ba95209"} +{"id": "train_40754", "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: 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": ["security_gates", "tooling", "auditability", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "798aedff363a2bf6a6821c1b1d7c0714d6d5d3ba"} +{"id": "train_40755", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fef9f8df837124b39a08dbc7ef5d3e0356a1cfa"} +{"id": "train_40756", "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: 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", "tooling", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfbc02050e4b5e4465da0ce3c6a427afcb4e9436"} +{"id": "train_40757", "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": ["security_gates", "reproducibility", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a18377aa252fd9b10874009b0dfb8dad207ece41"} +{"id": "train_40758", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "security_gates", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f45ebd61862149097c76b2b6410999b4a17c11ad"} +{"id": "train_40759", "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": ["tooling", "repo_scale_reasoning", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e9f3dbead09f276daa3c5a9bdfd1e3964f21d08"} +{"id": "train_40760", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c4157e4682befd30cc65dc1931b7d56eb8dd3bd"} +{"id": "train_40761", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a687c0e482a4393f4757362d2c60f599e0517c54"} +{"id": "train_40762", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f85f43a4bd67f27195adb2dda8b7798977fb7d7b"} +{"id": "train_40763", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "tooling", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c53f1ad644a42151b7f5d2e6bcdebe70635aaeb"} +{"id": "train_40764", "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: 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", "security_gates", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a28fd5c7c3b93d3d52dadcda67927a784ca52762"} +{"id": "train_40765", "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": ["documentation", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d20feb0cbd729e8de3ac332d3f13c7a82095f603"} +{"id": "train_40766", "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": ["documentation", "security_gates", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c92dac616e663a45d891970641ca850743411798"} +{"id": "train_40767", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "auditability", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80c9cd241f1e3e86653e5dd2f1599ecc7a47b1b2"} +{"id": "train_40768", "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: 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", "security_gates", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daed6a87b58cf573d556359cb2c0f5aa06ac335b"} +{"id": "train_40769", "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": ["auditability", "cost_latency_tradeoffs", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0eef96e56c254971a405401645dfa6abd1e132de"} +{"id": "train_40770", "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: 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", "tooling", "documentation", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9dbb9c0a0a3790af549124ebebb050295e992ca6"} +{"id": "train_40771", "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: 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": ["governance", "repo_scale_reasoning", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19f31fbcf20ba32d4481a0d5f9fda696f20e3e7d"} +{"id": "train_40772", "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: 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": ["tooling", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71de6b7eefdb4ab72a1178baf0ba6a2492b68af2"} +{"id": "train_40773", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a90c13373aedbe6efdfcbd34d90252f0aaa3cb99"} +{"id": "train_40774", "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": ["documentation", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "534c4f79562f554b90ad3e3e094b733664fa003e"} +{"id": "train_40775", "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": ["tooling", "tests_are_truth", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "942a7a28460223375d6da5bbce69bd29c8749c49"} +{"id": "train_40776", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3502d4c9de766d393ff4a309207ba372a19de305"} +{"id": "train_40777", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["repo_scale_reasoning", "auditability", "governance", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1e34154400abe6f945c7361863c72031ad2c033"} +{"id": "train_40778", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "ci_integration", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb24e5864c14ec346b4b099dac25f1eb273e66c6"} +{"id": "train_40779", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "810cfbd4582ed7a4ae03582a3de8ed147a4e0d2d"} +{"id": "train_40780", "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: 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", "reproducibility", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57cfeb3f8e1c37e6af897d62a954b1c711c5041b"} +{"id": "train_40781", "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": ["repo_scale_reasoning", "reproducibility", "auditability", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53d231d47c57d29cec5d43339cc59b8688b840ca"} +{"id": "train_40782", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7209e08cc9b2624c562cbb440382b69e8170fa71"} +{"id": "train_40783", "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": ["cost_latency_tradeoffs", "governance", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f1027d864ba6adc40b4b63aae28457cdef63384"} +{"id": "train_40784", "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": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd298fd15ba61c7dc8b967a256c9f4c31305c638"} +{"id": "train_40785", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4dc146e42fc87d8a6a4c8263673bdf9967422f1"} +{"id": "train_40786", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1efd76f690c7ef7de67e9b20387973f3696ec9d"} +{"id": "train_40787", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eebeb89bfaa6bfb01f6956937e2d8bb167bcd63a"} +{"id": "train_40788", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64a82012a6fca645fe87c92327e6ca3f334d8647"} +{"id": "train_40789", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c7ba8296b859eb1eaba854f4f5e2ca726ce66ca"} +{"id": "train_40790", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef11ad060b53bf36e25b4a9f029f8372143e08ef"} +{"id": "train_40791", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b03b88f3752dcd539f478018143fb38e5cb397d6"} +{"id": "train_40792", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f08ae9b4bf3a5eede7d292ee19c7ffd8478521b"} +{"id": "train_40793", "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: 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": ["ci_integration", "tooling", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a93153e7758d1b7242df3cc090267910fdbbad5c"} +{"id": "train_40794", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8063bfb35a69caa5a6307d2d59442c0cc69eb45"} +{"id": "train_40795", "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: 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", "security_gates", "reproducibility", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9cbbdaf87875ee733980529a66d25877e0535ac"} +{"id": "train_40796", "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: 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", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dd87bfaed1056bed87bb53b9556f74ef04ba67a"} +{"id": "train_40797", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["ci_integration", "cost_latency_tradeoffs", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ff89cca5ebe3fd80cf669d899dc5ffae2870e8"} +{"id": "train_40798", "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: 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", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11fb4d2aa91b0220ae2b78582d69cc272f67ce00"} +{"id": "train_40799", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d10558a132f162fa63e6b8fb8a1f0c1b2e2beb34"} +{"id": "train_40800", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d6b281c088180d455320e54b2290647f08a4157"} +{"id": "train_40801", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "reproducibility", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cec28592b5c864824d519c25c889a74ba9e3d8a"} +{"id": "train_40802", "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: 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": ["repo_scale_reasoning", "auditability", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e800cad1fe58feb6eff7ff26958f31041ca4823"} +{"id": "train_40803", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2fb2af9b00b0e74b6a37db50c69db42a34d1e0d"} +{"id": "train_40804", "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: 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": ["reproducibility", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daf7eab2e33aefc9b451752d59a94f267bd19347"} +{"id": "train_40805", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b81b8bc1dc7ad1e3b0685301a08cecaf86de39cb"} +{"id": "train_40806", "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: 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": ["reproducibility", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63893d19fa45a7e30a3ddbbfcc771325fc8ad030"} +{"id": "train_40807", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "864952cd658d46adccdbe43246317f95f7ccc6f6"} +{"id": "train_40808", "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": ["repo_scale_reasoning", "reproducibility", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1e4459347b20020552aaa7b57d836300b6e6c4c"} +{"id": "train_40809", "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": ["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", "security_gates", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fcac065030dfdb9990160e96bc2ab9a1e285320"} +{"id": "train_40810", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fe1dce00cb36c8fea47fbec75b3a7f0b4be656d"} +{"id": "train_40811", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "security_gates", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc4d6408c86b4598704e175595663bb423020a54"} +{"id": "train_40812", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1460b623022c7ff9f6a4c3794f25bddd3dacf43"} +{"id": "train_40813", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37bc0ca7add7798079b7052b1aa07209f39981dc"} +{"id": "train_40814", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e62ec4c7c101d1ce2ae2b01c2f806db59480abd4"} +{"id": "train_40815", "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: 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", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8994642a323e119c8f57c8eb9f6e3c6e66b5a97a"} +{"id": "train_40816", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "209c4cbfc75314b575a9dab31556f074de6bc51f"} +{"id": "train_40817", "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: 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", "reproducibility", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "541f5ca670135562715542f0c04df84f8fc54f39"} +{"id": "train_40818", "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: 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": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a76cd4910c13c4550660d795de0565c448b21388"} +{"id": "train_40819", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["tests_are_truth", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39ceb873ddcd250b3df54f7af3be934bff6cf816"} +{"id": "train_40820", "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: 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", "tooling", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d1ab89ebcd0a59e6a2e015b9f207ab71cf619d3"} +{"id": "train_40821", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tooling", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d8d76e17a71bfbbb44c4574fcb1d49de00e1a3e"} +{"id": "train_40822", "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": ["ci_integration", "reproducibility", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d60e7b9aedff94b2a7cf5afe3bb58e2f5c6ae9b2"} +{"id": "train_40823", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c6b6324552e978b798e26912a951bcbbcd44552"} +{"id": "train_40824", "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: 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": ["evaluation_metrics", "cost_latency_tradeoffs", "security_gates", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5831c6ff3aa3c3cad832aeefef3545e31ee5378"} +{"id": "train_40825", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3007ab9ee510fe60cba0fc8fa9922eebe255ec38"} +{"id": "train_40826", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eff70011e6b0c03b0e6403bbfad273b9350ffa5a"} +{"id": "train_40827", "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": ["evaluation_metrics", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed4dbf94e228f0243f0f90bf01a46f035c5c5dd5"} +{"id": "train_40828", "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": ["tooling", "auditability", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "072ceccf970e7af468a6fbafefa9a50665eceaf3"} +{"id": "train_40829", "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: 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": ["evaluation_metrics", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e886aa6bfa835f83005ea0565ba03dea7b3e1924"} +{"id": "train_40830", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1490b53e5c33d604aac137107724b96fc6ac72d"} +{"id": "train_40831", "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: 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": ["security_gates", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b18f1c4d212e72ca5b82b2aff030fb28e3dddefa"} +{"id": "train_40832", "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": ["documentation", "auditability", "security_gates", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f43a03d9395472d6064913f5fd289a849104c4ff"} +{"id": "train_40833", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "documentation", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5512c3ee665d279fd653dafed52bf21edc18cbe0"} +{"id": "train_40834", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42b15e3133690111e267af1ae7d9229c763912f2"} +{"id": "train_40835", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4121a1658cb75d5ed546151ff095703120447b01"} +{"id": "train_40836", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad1dce7f117e7400112f61d06c8a059ac5c98815"} +{"id": "train_40837", "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: 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", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f43b09a7926cc881dab988aa2f77576d68f647e"} +{"id": "train_40838", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9eea1f7877a03d1428d1af755e8eaf9fca13c26"} +{"id": "train_40839", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa7ab26228eb7a3186c2b8856d529e87499f4757"} +{"id": "train_40840", "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": ["security_gates", "cost_latency_tradeoffs", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24a9076a7cbbe5f7f892f4e0a67a1be37b1f569f"} +{"id": "train_40841", "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: 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", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf8ce836b38a4c74490cf8ba62e0a642ec643273"} +{"id": "train_40842", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12545497def44c1dfbc96ee218c0f48e53b3bafc"} +{"id": "train_40843", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "adce206ab043f2941a2a19b7b5541766e3e4a04a"} +{"id": "train_40844", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "security_gates", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb9f6facf4d80237b28caf2cbcf5fc49de080935"} +{"id": "train_40845", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11056e32456dbf44d1514d9a22e9cab18ac2ee49"} +{"id": "train_40846", "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: 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": ["documentation", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63d4a0be6b1cbe9b1247cacac1eaa71813d27cd4"} +{"id": "train_40847", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58d5fae0c5a0a36d155ab5141b7949d1341d752f"} +{"id": "train_40848", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b072d6bc9b3f850b8dfb0c58f7b6e9afe4bad031"} +{"id": "train_40849", "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: 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": ["repo_scale_reasoning", "tooling", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f438dd334d340ee89f935e6cc6415ad815d034c3"} +{"id": "train_40850", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57a62888b318138f4943f3f641447b8f6339f31d"} +{"id": "train_40851", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3a3f3b2ff97fa6f8d590d6bf8eb8ee0a6e17841"} +{"id": "train_40852", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c236ae2b18f19077279d2f58517f20bcf690876"} +{"id": "train_40853", "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: 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": ["ci_integration", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d055e06cb645861cd981b04e0f8c5480270ac1e1"} +{"id": "train_40854", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3c459fd81d885280826bbcca5587fc491269c18"} +{"id": "train_40855", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fef19b8770256373c822ed10319fa4035796ff06"} +{"id": "train_40856", "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: 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", "tooling", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25fcb1e1eac93b69fdb010c805a38f4409ce9d81"} +{"id": "train_40857", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a57904159b1f5d447aed044e843c744128ea2fb"} +{"id": "train_40858", "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: 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", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9bf6f3cedbffd8681ab7105b2201fce0fc102d1a"} +{"id": "train_40859", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e3943698dbdea7ba0e90509fddd142bd917885c"} +{"id": "train_40860", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b29d1a9ef8a9581c4117bf953a4350a53202342"} +{"id": "train_40861", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d22c282888083e14cc17e8499cc90d2b3137822e"} +{"id": "train_40862", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73418797b686ce3c9d7516843f91bf945604c1a6"} +{"id": "train_40863", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e497409dafda7d0c7977119e1846421e9f78817a"} +{"id": "train_40864", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85683775fa7c2b18c0dfd27a3148cb8cf591b3ad"} +{"id": "train_40865", "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: 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": ["reproducibility", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddd358b746accb46f54f7115ae201ff2fd215bdc"} +{"id": "train_40866", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53985bcf3895d022c5d4aca4b27648c447aca7b7"} +{"id": "train_40867", "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: 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", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cc67a923716501fa29f91a3b2eecd5f447d7a9c"} +{"id": "train_40868", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60f65e90b7990a89017dd7233e5d636e25255ec8"} +{"id": "train_40869", "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": ["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", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "609fb123da95bb02881c87d4f3fa32fc7373a3db"} +{"id": "train_40870", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c73f4a064f324dc82fc4c0fadbb4fecd5b80636c"} +{"id": "train_40871", "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": ["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", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "acd0d1647511dddc208479fc91eb3e61569d3c88"} +{"id": "train_40872", "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": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53e1d01d7d9396b05698f26035c6ab3c9584772d"} +{"id": "train_40873", "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": ["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": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0bd73c5637d677c55f591f8ce76ef041f82440d6"} +{"id": "train_40874", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "governance", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37a9d0f22d7879b19d020d1c430f9e5b1c46dc0e"} +{"id": "train_40875", "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: 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", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "402a63a5cef26541bcdada3d743711945920f100"} +{"id": "train_40876", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "evaluation_metrics", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "250fb929724e7b0fdc6659b65b4b3c314fbbf164"} +{"id": "train_40877", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "reproducibility", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "839967a06fc487bdcda80cb6f04e86904a7f16cd"} +{"id": "train_40878", "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: 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", "governance", "tests_are_truth", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cf51f9144480b8df8f98c94c2a7e1f44513dc01"} +{"id": "train_40879", "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: 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", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f43e1d43e980b6fc7776f50bc1c7b1390ee525e5"} +{"id": "train_40880", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9bfcd7675aad532cc4175b5a791fd88cdd83028"} +{"id": "train_40881", "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: 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", "documentation", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e3523a63e08ac5a6e3b520bde60af071db3f403"} +{"id": "train_40882", "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: 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", "security_gates", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c79da38d1d7a438e3d6428b2ab586a42a2dd1df6"} +{"id": "train_40883", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5349cf0b041786517121298701130e747bbb8999"} +{"id": "train_40884", "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": ["auditability", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf8a50e9b7571398c77f785e707ef1b71f834dbe"} +{"id": "train_40885", "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: 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", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c07ec15e4ac92435534c29e0533df1e6e37946eb"} +{"id": "train_40886", "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: 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": ["documentation", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35744582dd22a7ae645f23bb898b43168904c59d"} +{"id": "train_40887", "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: 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": ["repo_scale_reasoning", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5fc0b7a1315774fcc4996087c5f8433aed0bd389"} +{"id": "train_40888", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14d4813a8238321d56dbc44e2efff303f8192991"} +{"id": "train_40889", "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: 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", "governance", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93f562ed0b739f146dd447b35f13b47be9519366"} +{"id": "train_40890", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bc5c9159e76b5fbe7f7911902f5311d1b8791b9"} +{"id": "train_40891", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03edab330632860a9b8a9a5ba4cc951691fa9ff0"} +{"id": "train_40892", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f795b7ed3cba65c0f7fe84d7b4802b29d144ded0"} +{"id": "train_40893", "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": ["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", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63312aabfece2eb9a8d54e010c43bc67e379a88c"} +{"id": "train_40894", "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: 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": ["tests_are_truth", "reproducibility", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f232c51f20ae6d9ba65ab6aaf6d474a9fdbd7a0"} +{"id": "train_40895", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8be5c1359144136dd30c366c53b30abfffe0a337"} +{"id": "train_40896", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7faca7b599b86fe1214f29deef36f2e89ae0b409"} +{"id": "train_40897", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa5cd6947817c469c83e597488d5c12cca34040c"} +{"id": "train_40898", "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": ["security_gates", "tooling", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddff4f14c989a00c25bac6aaf2243bfb6ef74bcc"} +{"id": "train_40899", "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: 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", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1481876615ebac5bd17be3d742e5814854c05a3"} +{"id": "train_40900", "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": ["auditability", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0aefba1b07938e0fcf22acdfd256283306479aae"} +{"id": "train_40901", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db1dd8d5bcf6de357db15be76be225544e388f58"} +{"id": "train_40902", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8a0263b08b45603ed43110efc152665974ebb6a"} +{"id": "train_40903", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38ab4f9d6838203c5de30a08fe8b8afb40c6f744"} +{"id": "train_40904", "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": ["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", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76a46615dcc12079f79e617b651d1a619c0611a4"} +{"id": "train_40905", "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: 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": ["tooling", "ci_integration", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e6ef5cdbc3fcd23e38cf52774ae2125bd713b37"} +{"id": "train_40906", "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": ["cost_latency_tradeoffs", "tooling", "governance", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d7e13a2025432aba7d95cdabfffec4081f38987"} +{"id": "train_40907", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7209e08cc9b2624c562cbb440382b69e8170fa71"} +{"id": "train_40908", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "security_gates", "ci_integration", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03bdf5642fa83e14e8fa966001fadd9a6e29ea21"} +{"id": "train_40909", "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: 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": ["tooling", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37bff01c9311cbe3788fdf08bdde43f77f79c366"} +{"id": "train_40910", "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: 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": ["auditability", "security_gates", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "227b8f99fb9e51e475dd52a06af87be112b56f0e"} +{"id": "train_40911", "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: 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", "tooling", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dbd10b3711f73d65883db8efc794bb8b1299b86"} +{"id": "train_40912", "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: 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", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "647d333ad43a67a69ee3f0c43482bf4a7d75a56a"} +{"id": "train_40913", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca30049d907939e1bf2712417502d9725b1265a6"} +{"id": "train_40914", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08a121f87f85a2f0658eb1d31d52de83ce6312ab"} +{"id": "train_40915", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc989d34743ac5092cf0c1105fe5b15720947e73"} +{"id": "train_40916", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cd88ac7f34566db185060201c5d66bc7820b939"} +{"id": "train_40917", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "555b67e35006f97b24b5aabc982a1e65262fc047"} +{"id": "train_40918", "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", "documentation", "auditability", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ebf46b1d260d131f4edab8436b04855abb14c85"} +{"id": "train_40919", "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: 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", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efb37e37f34f68c277b6c42e4d0e2ab9a65ff216"} +{"id": "train_40920", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7543a8ae8973a57de726d90ae36f763a5797de8c"} +{"id": "train_40921", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "571a90bea95767bf0625da74fa1d1f42f42900ec"} +{"id": "train_40922", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8da6080d73f1970c545a5968b4f6663c3b09c8df"} +{"id": "train_40923", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "repo_scale_reasoning", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb52feac4a908e7a7bf9e06dd085f7df651f262d"} +{"id": "train_40924", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "auditability", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f555c4e34807e78316cc6eac2ec78d88d3c9ecc5"} +{"id": "train_40925", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "documentation", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c8dcaa34c83ecc467bf73762db2cf6343d0a5aa"} +{"id": "train_40926", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ca091162c1b9a06c01b99fba5e1a7e3a0b27aca"} +{"id": "train_40927", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1b71de7cc98980b6bf8adcffcbe2fcc8367ba53"} +{"id": "train_40928", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["auditability", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c635e09824c880587f1556789b2fbe4a7565ba2"} +{"id": "train_40929", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1e48c8766996b5fc2865f207af4bed3ace4b587"} +{"id": "train_40930", "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: 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", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af6fac4b9e7274cb6b65913a741a7c029af49403"} +{"id": "train_40931", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac93e74ee9c0e4c9ede2ef5c78a9732faf979f9f"} +{"id": "train_40932", "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: 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", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8be278f72e10a7ca7c85c0f8ef27b901269f8c1"} +{"id": "train_40933", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0df60caf39aa452f6c92b2381cb51c13cbed97ff"} +{"id": "train_40934", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67a071be227e45399dcb098d9539d7633d5e69d5"} +{"id": "train_40935", "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": ["documentation", "tests_are_truth", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c5ca8edd44b03c64491d722a85df4923f5cf7bf"} +{"id": "train_40936", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9baad2b01aace3a8304b65f3c919e31cc0ea06e"} +{"id": "train_40937", "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": ["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", "cost_latency_tradeoffs", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2c5a8d80215fce9c7b4ac485731799539a561e3"} +{"id": "train_40938", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37ba3767890ffb9c6c5409240ee56fb212e38f2f"} +{"id": "train_40939", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f6fc01269f0bc9bcccc117274799399143b9de2"} +{"id": "train_40940", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ba5f8cd42a5f1bea8058919b71c1b47b428579a"} +{"id": "train_40941", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3bd8ec65917d76745d3931754fb6f48e341a966"} +{"id": "train_40942", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24cbe6e22ef86e0d07bb00c2f0d60e126330372a"} +{"id": "train_40943", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e143d352b973ff15bf59858f068ffd0040501110"} +{"id": "train_40944", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready 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": ["ci_integration", "reproducibility", "tooling", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ffa4e0c54fd5c49b73ddd268a2def2d87420e22"} +{"id": "train_40945", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a289ab41451c41eebbda414df5799195f4e3cb04"} +{"id": "train_40946", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "security_gates", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79b34b466c8ce71a507d14c161161f29ba466d6a"} +{"id": "train_40947", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ce23e54c276bc7aec49fe40f4ae61bc298c9d7f"} +{"id": "train_40948", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20eae69fccc88d7fed4961de78ebfa17b15adf0b"} +{"id": "train_40949", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "031699eb7a14486a1bd018dcc01e04af61ce0397"} +{"id": "train_40950", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "reproducibility", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2365b4b9770fdfd1ba8325cea7f6248c2fc9eba"} +{"id": "train_40951", "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: 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", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0813654225443d0dedd55e53cdc070ca89dfc2c9"} +{"id": "train_40952", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9b2e23698f62bdf4cfdc18364dce65c8b4f2574"} +{"id": "train_40953", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4da685f40083326807e0eebd3a9b3a4a9ccd2cd"} +{"id": "train_40954", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e3a5c915f0791fef0cf33d9b551291508381345"} +{"id": "train_40955", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb769ad27bbf7b27b918f24072815ae69924ca63"} +{"id": "train_40956", "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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64d47272dec5dbadd197ad3df2ea7084f50f1193"} +{"id": "train_40957", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5df37db7bb59ce6137ed7f94026de7466cdaa0c"} +{"id": "train_40958", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f00a85d5180da91ccc97d7cab8b53f15987acb80"} +{"id": "train_40959", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "379a5a36060b134454624db7f03f7b1d13f3bd77"} +{"id": "train_40960", "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: 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", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ee0d4e6e41056bf6101881a21759fba51ad156d"} +{"id": "train_40961", "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: 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", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "638a8870021ea18900e93c346cd0104a21eb184b"} +{"id": "train_40962", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "479d3cbe61db2c1b42f6efbe2507c74849476e23"} +{"id": "train_40963", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "caa1417806654445b9216a6dbdf459e6300f6621"} +{"id": "train_40964", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf99a3f20beb7b56eb3dce6ebd4848db7458f0ec"} +{"id": "train_40965", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["tests_are_truth", "repo_scale_reasoning", "auditability", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c1f0769ce3b4b15d1db1a939295bc63219e7b86"} +{"id": "train_40966", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e9e99679dc32584ae0605fe023e63d52bffcd01"} +{"id": "train_40967", "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: 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": ["tests_are_truth", "tooling", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c66fa4110e451e61116258de3b960a865e0cf3c"} +{"id": "train_40968", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3eb2303a4270f991a3ce911829b0954677cb50e"} +{"id": "train_40969", "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: 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", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "929bd47d24d00821a12858b4fa74103657f4b922"} +{"id": "train_40970", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "security_gates", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b64891b0482db718510eec8a8c956978cc297f7"} +{"id": "train_40971", "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: 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": ["reproducibility", "cost_latency_tradeoffs", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dae5f6bb99517deba42f6a88f668b02c19166bf"} +{"id": "train_40972", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "928c3e27521972271934655e63b25b8bb44aeb9b"} +{"id": "train_40973", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29f43ccef72af2dddbc6f4e132bab073c5d648e7"} +{"id": "train_40974", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tooling", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e34b639bcb13001c0cc23ed4807046d0bec98433"} +{"id": "train_40975", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a46e7aa438f06bc2ba9458355a88ab061f9327f"} +{"id": "train_40976", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "governance", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f555c4e34807e78316cc6eac2ec78d88d3c9ecc5"} +{"id": "train_40977", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2444bd9ecf0f0cfba02e6ba4c10cd745b0193f7"} +{"id": "train_40978", "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", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f032e0ef0e2fd0dc053d8bf740342137b7799dd"} +{"id": "train_40979", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30e8a21ec10839b32a979bd9d2052ebd69e41450"} +{"id": "train_40980", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f622d4fb33642ce0836af29421b1ea7df5cc6de"} +{"id": "train_40981", "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: 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": ["documentation", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbd9ad7491029c94ae1afccd9156c3dfe68e57aa"} +{"id": "train_40982", "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": ["reproducibility", "evaluation_metrics", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb52feac4a908e7a7bf9e06dd085f7df651f262d"} +{"id": "train_40983", "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": ["cost_latency_tradeoffs", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53247b8a7fff1b6c0790f2e76d419f7333cceeb0"} +{"id": "train_40984", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc1a7a6093780c6ce6b826cf7f99bbf4eabc2d30"} +{"id": "train_40985", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96a1cc034c772f484b13e6552b7481c17b1d5f33"} +{"id": "train_40986", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["repo_scale_reasoning", "security_gates", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e41738627729729a37441c013bde4ac182d0778"} +{"id": "train_40987", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17dbd1e5d118c0dda985909f156919afbfdc0f5d"} +{"id": "train_40988", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e99e5c3defcdbc5a187a3ad423e5528ce1b10167"} +{"id": "train_40989", "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": ["documentation", "ci_integration", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2750524268a79692ffe21dd8880276942ac8d15f"} +{"id": "train_40990", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60d1d530396e6904e18bd8b2a8ce5554fd89dd30"} +{"id": "train_40991", "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": ["ci_integration", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d39e2cb1a32b3229c877982178378cc7c1a011eb"} +{"id": "train_40992", "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: 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", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85e1c393e4c6a23397d06bf5d4f30bd3e8fa09a3"} +{"id": "train_40993", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd8ea54936c77bce07b9bd86ee0dc75b93a6a056"} +{"id": "train_40994", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17e269b3a33a448236f9b9e0c2a79081f8104ae9"} +{"id": "train_40995", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98905552fa64ed883f10dae5376ac86485af1fc0"} +{"id": "train_40996", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0a444cfc1612802ad5100a8b0e5e5270509be7b"} +{"id": "train_40997", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13a6cb0a7a8ba5e0d6cae9d2675a99d621978212"} +{"id": "train_40998", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b424b42b1b2fda4c9bc801a8548a6e323b06f03"} +{"id": "train_40999", "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: 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", "governance", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "986fa607ac2986c771a312615e1ee3a734745917"} +{"id": "train_41000", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b472e7ba8c596e5b4e5385b07cc47ed2a528ab7e"} +{"id": "train_41001", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f89ea9296005fd51b1ca0f108cf07ccd0bc69df"} +{"id": "train_41002", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51b6047d4c3e589837311f710f9f6f8cbb3b3c4f"} +{"id": "train_41003", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0e4054eb016d294c90b9a5ee2a24332db80fdcc"} +{"id": "train_41004", "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: 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", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a708435587bddf5a806cbd9bbe0402e6511e5e0"} +{"id": "train_41005", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "060af3bc76b5a27ba780357a7586e95c679eaa42"} +{"id": "train_41006", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3add3d6fb41910e951d117dbd74ef92a33a1041f"} +{"id": "train_41007", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95c7ed704c16951b82bcc0b1f58e43ee0455e065"} +{"id": "train_41008", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a89be84e8d6e6b59c5546c9fa97bf7797ef64de"} +{"id": "train_41009", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8d95127cee3fec536e93b23837fce48f8b50001"} +{"id": "train_41010", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "documentation", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71dabc92bba7dbf6a3c1d0fc8c32e43385aa5f50"} +{"id": "train_41011", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9dcf720f523e9a761e0c813d5a9ab087e78158ae"} +{"id": "train_41012", "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": ["ci_integration", "auditability", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcdd1cf58908279001c7214be5c47352dc58b0ff"} +{"id": "train_41013", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0bd732078567af9978a7f98e12e6fab2e58eb3da"} +{"id": "train_41014", "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": ["auditability", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8abf18769a90a1ee11138e46f6d20f8291ce245"} +{"id": "train_41015", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ceaea7dca75451790b9d650109ed26ddce17c153"} +{"id": "train_41016", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7f930fdee283fd57890f75bd13384438ae81f2c"} +{"id": "train_41017", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7a98e74e04ba87ac535f6d318879622ad18e147"} +{"id": "train_41018", "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: 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", "governance", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9f3f4c05dfa065f3336893f30a2e205b1ea9eaa"} +{"id": "train_41019", "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: 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": ["governance", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f43b09a7926cc881dab988aa2f77576d68f647e"} +{"id": "train_41020", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "tooling", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10eb1962f692fb36c12c86d774fa9aca2df06e0e"} +{"id": "train_41021", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "tooling", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fc22a4c3babf05088f28f28ad07d9657249d321"} +{"id": "train_41022", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "693f05be8dbc2ca492ff44f3a4992d72614b12ef"} +{"id": "train_41023", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4da685f40083326807e0eebd3a9b3a4a9ccd2cd"} +{"id": "train_41024", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2596f136332ab9e4198ce64b24020a62f0ae0e1b"} +{"id": "train_41025", "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": ["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", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3166df9a678c95bfbe651947d3e9ec74865d1516"} +{"id": "train_41026", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5ff58f871f6f73dbf6750234a5fce9c6cb0fc85"} +{"id": "train_41027", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a66f708957a4d41625dfe0c16f0f627d1227c2e6"} +{"id": "train_41028", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3dbeb7fac82ddcd6701f441faf6a385d4a3cd95"} +{"id": "train_41029", "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: 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": ["tooling", "governance", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "764e89c29b808941da076134835c0f097ea23820"} +{"id": "train_41030", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d483d60aac07b7491092a9cb4efbaae53316cf21"} +{"id": "train_41031", "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: 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": ["auditability", "reproducibility", "tests_are_truth", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "962a953e5c472cdad4f545026cc297d7dabff220"} +{"id": "train_41032", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9452c9657db433ed982d422e4fcd1cd91b3bf541"} +{"id": "train_41033", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f2a0950656faad5cb273ef5677ed1a6644e89af"} +{"id": "train_41034", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb50ef30768ec9a3c3032d2d1ea5e4f92b336e5f"} +{"id": "train_41035", "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: 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", "tooling", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e32ebb56971c489dfd62580260f16f4ea0771ef6"} +{"id": "train_41036", "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: 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", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f00e9cf33dbd9c56e30b63b2f171f1df7607ab77"} +{"id": "train_41037", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "governance", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f46c44527e18fbcbc2e7329ab64186baf28c8e39"} +{"id": "train_41038", "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": ["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", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5e3edd3267edeef928d032dd3110e3976531e8e"} +{"id": "train_41039", "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: 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", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5000b0d1ad1eae8bec45deb71b93c0e9f21c5ae3"} +{"id": "train_41040", "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: 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", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fc3c9d2549d1c5a4c24e20e5ac9b4de0b50a61c"} +{"id": "train_41041", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebf5f532383cf882e5dbd7100153384dfd7e02a0"} +{"id": "train_41042", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e94d7b95a8b8ea610e050f9ed2713da6644a70df"} +{"id": "train_41043", "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": ["tests_are_truth", "tooling", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e91c4d723eb1a21682c6fb975724c83cd9281b2"} +{"id": "train_41044", "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: 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", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62162af32264eb20e045cd71aa7f34c1caa8b039"} +{"id": "train_41045", "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: 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": ["ci_integration", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d5a558b11db5bb85c7867ac48f7011b7aeb3b3a"} +{"id": "train_41046", "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: 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": ["tooling", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85ff536d3f5b2bbcfce3796b57596b5c50684016"} +{"id": "train_41047", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d0fd679a659f1a21076e7f24160976f2aafe9f1"} +{"id": "train_41048", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["tooling", "documentation", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f65438c8752a56a060ba45c16c80fe231c15ecb"} +{"id": "train_41049", "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: 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", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3c6b7da98dca70522fdf4fcad328ef9b74ed32a"} +{"id": "train_41050", "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: 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": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48932a918397d17b10a13bf1f8485c280964143e"} +{"id": "train_41051", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "084a058e646ecf5641df2eed92a42e18c18bfeb5"} +{"id": "train_41052", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60b5dc83818d4650ad54f086946eb4e904f30a6b"} +{"id": "train_41053", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tests_are_truth", "governance", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5de64494073615e04760d2e19637814d239fde7e"} +{"id": "train_41054", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "documentation", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efc4d3089e5e1681345e793d573329314fd8ad0f"} +{"id": "train_41055", "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": ["ci_integration", "governance", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6108e9387fb129dd6f501d63faece6064d2074d7"} +{"id": "train_41056", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6402e6d77010e01cb86241afb67a45bd195f77e"} +{"id": "train_41057", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0445b6e65a40198d41b897a1fa23d32a8b74d511"} +{"id": "train_41058", "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: 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", "evaluation_metrics", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bf35b16231c975f8fb41a865a14ba4bf95be793"} +{"id": "train_41059", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2930813a4d6c7a2263ebc918fe23c8db001e6052"} +{"id": "train_41060", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e96ac32fde18709fe30eb74f39a87c68f6de4aa"} +{"id": "train_41061", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["security_gates", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8467c04bd3693b52c89b60f91f8d4ec0da6abd8"} +{"id": "train_41062", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["auditability", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9074539fa9a0aee5978d8bca9b55362e4841c3aa"} +{"id": "train_41063", "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: 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", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1eb89efa4c56171a7ce61b9043bef3ec0d9ed0b5"} +{"id": "train_41064", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "035d4df7a0d1380b9ed3a5abaa3b9800149b5aef"} +{"id": "train_41065", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "reproducibility", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d008a2e469279df12a991eca8e30a8b6c28d40f1"} +{"id": "train_41066", "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: 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", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33f60a2c4aaaf6e0edcf53548bd52bf7c084c56d"} +{"id": "train_41067", "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: 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": ["cost_latency_tradeoffs", "auditability", "governance", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a2dbcf43c1db30c7bc6db3238744a66c31e7093"} +{"id": "train_41068", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78617379256ee0946e31044c86be4a6a4beb80ff"} +{"id": "train_41069", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "352fc7623662d2f5d300746eaead465d9a92572d"} +{"id": "train_41070", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8237d8848555c80ca12f4062b4953612abfa3d54"} +{"id": "train_41071", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0962e4cd11a2af01aa14162ce958bf3a779ae81f"} +{"id": "train_41072", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4593362a73708ec16c6a82361280aff54c0920fa"} +{"id": "train_41073", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c767c4a69e50d80568e58468079364d10bd62a60"} +{"id": "train_41074", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "security_gates", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8175dec1eee85b881365fccdbe3043878432297"} +{"id": "train_41075", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1fc008225378da75c5a7c96ae08b42ddd4fdae5"} +{"id": "train_41076", "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": ["tooling", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1c926ea1a1fdc13feb9fbbe3d42d4289578efc1"} +{"id": "train_41077", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ce99cdfd7fc168f5e8f0119791527ed92f5dc87"} +{"id": "train_41078", "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: 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", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d24f667733703eda31121da3d84102133d96fa1a"} +{"id": "train_41079", "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: 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": ["documentation", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e24030bb0e7cd624d3cf8091ea0101e906efd01"} +{"id": "train_41080", "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: 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", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0d0ce76b9b68cb2e05ba37bf71fa957c8c19d40"} +{"id": "train_41081", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8830313816679601f3e405e469cfe5fcde35e5b7"} +{"id": "train_41082", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcf6ef9260dee6c58e9378b177472f60d9b49a85"} +{"id": "train_41083", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "documentation", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eccb3607745cbe4f0c3081f0ab0d30ba6c87ecb1"} +{"id": "train_41084", "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: 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", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa0d7de2d5002ef06a46cd1f5c5d4ff21cd23807"} +{"id": "train_41085", "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: 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": ["tooling", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0422a27f23fa4e930664b0cb217715315e05fc54"} +{"id": "train_41086", "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": ["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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9768452fbb74886295779a4a8f0ba320fd2aeea"} +{"id": "train_41087", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "445b3b58988154c1ff762133e203b3a3141d5dfa"} +{"id": "train_41088", "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: 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": ["reproducibility", "security_gates", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52c4a0cfe004b0e224e46837921f98a5dd4f0f9f"} +{"id": "train_41089", "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": ["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", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "862150dcf38eb6803f764de0cae2f7f1ba1dfc4f"} +{"id": "train_41090", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "400fb50998c0aed0ada515fe8dd79d6fb57071b9"} +{"id": "train_41091", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d100abd0df2c9d8fe65649dad696d5231aa376e"} +{"id": "train_41092", "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": ["security_gates", "documentation", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb10fb0afc24df06b713e6ffca90120bcee30440"} +{"id": "train_41093", "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", "ci_integration", "governance", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fd2094d92a38c737918753709563a2c97ab0032"} +{"id": "train_41094", "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: 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": ["tooling", "ci_integration", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1710063d3f9663dc06c29726e5601270b40df702"} +{"id": "train_41095", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7244ef29260d2dbb7f876dcf83a44e5b593cef55"} +{"id": "train_41096", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "241d6c23a53c00620f3b75767339440eb7254c48"} +{"id": "train_41097", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c1f0769ce3b4b15d1db1a939295bc63219e7b86"} +{"id": "train_41098", "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": ["reproducibility", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ec5d26aadb0539b2ba3f3304158315418d2ccf4"} +{"id": "train_41099", "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": ["documentation", "tooling", "ci_integration", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ed7fadb6e2ef2e72b00db96ff46729ac9c6cf3a"} +{"id": "train_41100", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "cost_latency_tradeoffs", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfc6f4aa4ba246a63b5206e06a1465f51744a6af"} +{"id": "train_41101", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4917d681291af8f8a00f19f7fbcaf617c99176f6"} +{"id": "train_41102", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["repo_scale_reasoning", "governance", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33e14e93eac888c009df11487ffeb51bd98ca281"} +{"id": "train_41103", "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: 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": ["ci_integration", "security_gates", "auditability", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1b4e0f14a0285341597f41601fefa52748f5ebb"} +{"id": "train_41104", "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": ["tests_are_truth", "tooling", "documentation", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "654e23ec3772c9f2459273b4adc9ca70ebc0220f"} +{"id": "train_41105", "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": ["documentation", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a98a4a917cae83fd3280c5e4c35f464b6518b3c8"} +{"id": "train_41106", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e10ca851f24eb3468bb0eb8fd2903538741b7cf"} +{"id": "train_41107", "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: 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", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b75bb9c880be4e0006d0936707f2cea53d017341"} +{"id": "train_41108", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["tests_are_truth", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4748b13fbf62cbf5267e507a7914634c92cdd538"} +{"id": "train_41109", "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": ["repo_scale_reasoning", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c071297246d610823e91dc2a5ff15bb6ab99832"} +{"id": "train_41110", "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: 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": ["documentation", "auditability", "governance", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a24f6533de505f92178b7a36f2d1f985e3a4e1a"} +{"id": "train_41111", "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": ["tooling", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e498ae8b098874941974cff60ae1b3fff94bd02e"} +{"id": "train_41112", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a7ffa22e6a2b99e969419812f47dd169cae323d"} +{"id": "train_41113", "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": ["reproducibility", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f43a03d9395472d6064913f5fd289a849104c4ff"} +{"id": "train_41114", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f37c48ba3eeb92b8f9f5aca47310930819edafec"} +{"id": "train_41115", "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: 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": ["reproducibility", "tests_are_truth", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddb75b7545a1c1ebc21f2f80adb5e3b4f4b5b7de"} +{"id": "train_41116", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd2555e5d657c5f50830294b2a2594ed4c25ac32"} +{"id": "train_41117", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcfaf0e2b34977a8a181901e3104e3bfdab18656"} +{"id": "train_41118", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "governance", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4db46c163bec03d649d607081f56c7aef5849f47"} +{"id": "train_41119", "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": ["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", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ab162663d2b7af924c71f82d15d66bc10c7e743"} +{"id": "train_41120", "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: 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": ["evaluation_metrics", "documentation", "tooling", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64d51daedd84faec68ba3ad512967d3daef2bcee"} +{"id": "train_41121", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2dde083a93a0bc415bfbde90b225214adf0086a9"} +{"id": "train_41122", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17269a6f04a4f03fd64fc7e91eb704d863c8963b"} +{"id": "train_41123", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d881b93b9c6bdf00bb58eb39918ecc368f4c8b95"} +{"id": "train_41124", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e88ea99e21fdd479dba6d1f1d41827c03347743"} +{"id": "train_41125", "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: 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", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1465ffcc00e839f0f0720245869eaa4f119341f7"} +{"id": "train_41126", "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: 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", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "727dd238e03a35d7233194c480eefdc10bf317c3"} +{"id": "train_41127", "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", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b396d7d5566a2f9b9e9169760e328b2ff976bca"} +{"id": "train_41128", "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: 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", "tooling", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9596786df1ef197704ff0b2a28015230bbf6b437"} +{"id": "train_41129", "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": ["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", "documentation", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cefdc68e2306cfae5718a1a04c1cd278484b296d"} +{"id": "train_41130", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "C#", "developer_needs": ["governance", "ci_integration", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42fe14bb145bbde31a33af4b3cbbed84c26a05db"} +{"id": "train_41131", "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: 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", "security_gates", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f8991f6510d681425ac31277cef23a37e1d6a94"} +{"id": "train_41132", "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: 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", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d25babcf3f47f26e78a40edff362fbeee8b2c319"} +{"id": "train_41133", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0ee6400e225d6e44c660300ec2e242230418281"} +{"id": "train_41134", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "343394c7533375e2a57a6a4c8c1df5be42a675a7"} +{"id": "train_41135", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "266e8e933ed9562091906793501d6e217d8b209b"} +{"id": "train_41136", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b720c012383d00a7d0fa37176a85144d2c441f6a"} +{"id": "train_41137", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecf43aab777ca4347221fae797bd13d58fac9dac"} +{"id": "train_41138", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13a88ab0a1daa1a5501b369f0189fd446fdc5707"} +{"id": "train_41139", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "governance", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c52b02fdc1be448f4ef01a0e75ab8ca5a3d5d14b"} +{"id": "train_41140", "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": ["tooling", "auditability", "documentation", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19c8ae1aa49b78b86f265f1d590f7be508a83c4a"} +{"id": "train_41141", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41b701e43b6267f3f20d18d8f5a66082b852edd2"} +{"id": "train_41142", "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: 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": ["tooling", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44c6dca66302d5dcebff4b585cbed5009bf50440"} +{"id": "train_41143", "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": ["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", "auditability", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c5cc823c6893235f796cc8a06cbe9e477912310"} +{"id": "train_41144", "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", "documentation", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6622daca6567d2a3dd54ac54ae56000f25334493"} +{"id": "train_41145", "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", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "987f64161d352c727afd814d8878c7eb0592fe1e"} +{"id": "train_41146", "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": ["security_gates", "reproducibility", "tooling", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dd44274fa2df2f9a5b3559b088fbf7d2faf5552"} +{"id": "train_41147", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8ac04baacc156327329e2022e2d424abdc0868a"} +{"id": "train_41148", "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: 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": ["evaluation_metrics", "ci_integration", "tooling", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "101fb5120da96a32a939af21c75dd8e8c6217ba4"} +{"id": "train_41149", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "documentation", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d599f21f05562a0154c54f1add15d8f527ae2056"} +{"id": "train_41150", "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: 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": ["tooling", "reproducibility", "governance", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83eebc8ef4b6bba4231e92b7d959590e2c514a30"} +{"id": "train_41151", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20a9a9d47d85b9e811990e2d3a79b336b410a980"} +{"id": "train_41152", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["security_gates", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "076a6891bc2fb79069a57465291336d169ce75c0"} +{"id": "train_41153", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c88f96d4a4f4724ef3a1a5fd285058e255b5057"} +{"id": "train_41154", "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: 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": ["reproducibility", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2706d69657bc71093ee48051c507349aaf2a40c8"} +{"id": "train_41155", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3eb95e1e9f9737441e613088b303c8b2b1486580"} +{"id": "train_41156", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "tooling", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5ce68aace23aec509eaf1079d235ce993ebb37f"} +{"id": "train_41157", "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": ["tests_are_truth", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5f9f4c76a9d8af65b4075e58b78a9c979b4aa61"} +{"id": "train_41158", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae6d654d884a394398b11b2778ea7c688c9f652a"} +{"id": "train_41159", "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: 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": ["cost_latency_tradeoffs", "security_gates", "governance", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54fa13c0d817d38ba5c58eb152a9459bcdbe4073"} +{"id": "train_41160", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f6b2da27cde63a8d5076211c3b671526b2daa32"} +{"id": "train_41161", "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: 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": ["documentation", "repo_scale_reasoning", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bbd9757c3becda474812f237c5623d8fd742a7f"} +{"id": "train_41162", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d691551df51894e5d07829220e97b2338fe2aad"} +{"id": "train_41163", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8b64f0ceb55734aede28dc48bb15df7540b7bac"} +{"id": "train_41164", "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": ["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", "documentation", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c06a54422052f130431c8428bc3c381085a6d7f"} +{"id": "train_41165", "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": ["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", "documentation", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e50eb6a619704b7ee6fe5037ea59a0fce37c3d9"} +{"id": "train_41166", "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": ["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", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ce9b0b2e70262a0c702c4b304da793e5615760e"} +{"id": "train_41167", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55959b1d15f5f7689c30e4c1df93a9932743c916"} +{"id": "train_41168", "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: 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": ["governance", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c217de961c2f09687a2759b705db1ed5b4fd15cf"} +{"id": "train_41169", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e8ffb45142b41376e3bc3f796ca563683df5e65"} +{"id": "train_41170", "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: 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", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25885baa095720e7869db070d383567ef6853286"} +{"id": "train_41171", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b532327361221961dbb90b46ba2a679067b8eff"} +{"id": "train_41172", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d40aa165df614e9927eac3738c30c7036d26b27"} +{"id": "train_41173", "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: 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": ["evaluation_metrics", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ec2600af037c7f4460a1e436cc2db76855c14c8"} +{"id": "train_41174", "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": ["reproducibility", "cost_latency_tradeoffs", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fbb2f13a0aa8b2e5303e0b3a8f2133d7a91178e5"} +{"id": "train_41175", "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": ["documentation", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22f42f8c834e48a87b24b981d625f0b5e1f82e98"} +{"id": "train_41176", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2760483a3ba592fb2b88eec49cab627fba146e96"} +{"id": "train_41177", "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: 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": ["tests_are_truth", "auditability", "documentation", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ac52b426ee410e3b6af2bc5c342040906eac227"} +{"id": "train_41178", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6dd67450c48ee6179a3d02ec3b2d1a6964b2862"} +{"id": "train_41179", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8f1d8c99ef434489aa80ea509c27b005fd7d10c"} +{"id": "train_41180", "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": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe887e6cae86df26ed64d90fe33358860f870b5d"} +{"id": "train_41181", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa89f9031ed37e4fd5ede39d21d4c4e5cb011a56"} +{"id": "train_41182", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e37d21244c01690b8c7382071718eccfa4ede756"} +{"id": "train_41183", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["security_gates", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2512fca105cf411028a33988ee3c1184ae478d51"} +{"id": "train_41184", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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": ["cost_latency_tradeoffs", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa377164f550c87214e5cea8a7f66d0e7c2d42b8"} +{"id": "train_41185", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4777561d786acb2110eb91f4cf1d30cb7a84fb5f"} +{"id": "train_41186", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "332e1f5dfae3f544b266d1b815d9f2a0d786886c"} +{"id": "train_41187", "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: 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": ["security_gates", "evaluation_metrics", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9584ef0b3da2e07c2a73f8eec851066353a57acb"} +{"id": "train_41188", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "documentation", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2011255debaa72bf89a40ef24d348d02cf61a01"} +{"id": "train_41189", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c72abdb2f98fa3cbad5af5592045103f9b6e0f7"} +{"id": "train_41190", "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: 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": ["repo_scale_reasoning", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef09f514935d38e29124b17de287ed2e06cf95a"} +{"id": "train_41191", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8eae8c4e55fcdb5dfc4a681d1053b1f739900aa"} +{"id": "train_41192", "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: 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": ["reproducibility", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2629ca7f2a1bf9ebe1d0116f5d5e5cfe3cffa534"} +{"id": "train_41193", "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": ["auditability", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ac128bd1a493b3b76afbc509dca25a4ab36b449"} +{"id": "train_41194", "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": ["governance", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "976d21d3ee75ce7f17c61255223052609f0c4326"} +{"id": "train_41195", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6c145cef6418ee8f71c83374951789b20d8f380"} +{"id": "train_41196", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56213a35dbfacb2127fd6320e64ebc80ee22912d"} +{"id": "train_41197", "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: 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", "auditability", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0422a27f23fa4e930664b0cb217715315e05fc54"} +{"id": "train_41198", "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: 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": ["auditability", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bcd3740aea2cce7e2004289ff1f44752a06c277"} +{"id": "train_41199", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "496212de7fa3c792e95c60281d79e30597f6b556"} +{"id": "train_41200", "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: 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": ["cost_latency_tradeoffs", "documentation", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52c4a0cfe004b0e224e46837921f98a5dd4f0f9f"} +{"id": "train_41201", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07b9c501e82548571c53a2ce8eafa22f6e042866"} +{"id": "train_41202", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5e4d6dd7f61fbacee0391db6189fd0d7ac81877"} +{"id": "train_41203", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5b1b6c8afa5a95f410cc5051090b4b2b6130536"} +{"id": "train_41204", "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": ["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", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e16b164e9c39f9f4b8279e86979abe0bd3ced9e5"} +{"id": "train_41205", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["tooling", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e79dac0a6946fb66030be12295e3fde044cea1a"} +{"id": "train_41206", "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: 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": ["cost_latency_tradeoffs", "governance", "documentation", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a574c415142b061c08024e6738167c853cf5c327"} +{"id": "train_41207", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6b044c188e8f2bc73eca523261ef42b2f358374"} +{"id": "train_41208", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34910bbc31a666a9e7023d780bbee03f1157b2ff"} +{"id": "train_41209", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d59addf2c18f6619ea8c150d07e4a8d59a9de049"} +{"id": "train_41210", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9405dd5ae8d30d29422876af487de027c656e6ac"} +{"id": "train_41211", "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: 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", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a196df52043bd4e72a9aa0768039e879b1318dd"} +{"id": "train_41212", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac523aa8af9cd5e753cac9b4ae5180128a5bb647"} +{"id": "train_41213", "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: 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": ["ci_integration", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1908c7c21f0021f41e6294f919c6aefad7c10095"} +{"id": "train_41214", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f51f36c7873d906bf2e6fea05cc29617c7c179c3"} +{"id": "train_41215", "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: 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", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27d0daf20cc1ce816a4e1cd541b2f6da95873bbd"} +{"id": "train_41216", "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": ["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", "security_gates", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed312d31d2955b82a86d2cda769a08c499772bfd"} +{"id": "train_41217", "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": ["security_gates", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd6843281ee952a156ffe34cb9de205d66a5ed76"} +{"id": "train_41218", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bf1d0513d4d885d1ed5f7e05a214f59b99efe02"} +{"id": "train_41219", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "589b6fcec6d666ed06307f2db6cee70ccc3c140a"} +{"id": "train_41220", "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": ["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", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7ed5cf9c0c084f399d7eafc9749d5de38ce321c"} +{"id": "train_41221", "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": ["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", "security_gates", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "edcd5fe162b95ff34b5d2154b9e53581491c6817"} +{"id": "train_41222", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a62f5f7959731cdbc19f1d23bb1de06a47fd8be3"} +{"id": "train_41223", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "229f7323896feb43e8e49103636132f6bf44288b"} +{"id": "train_41224", "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": ["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", "reproducibility", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "804d4cb051d59705cd9a802898e22e8f13407ed9"} +{"id": "train_41225", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["repo_scale_reasoning", "auditability", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ee1e4e784c6a400d44c6852d408d05cc84a0760"} +{"id": "train_41226", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f190231fa22d78e844c94f14fdd2622752b41cbd"} +{"id": "train_41227", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ea4e1b71082ea0b552971398c08ac6cb4075d8f"} +{"id": "train_41228", "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": ["repo_scale_reasoning", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2777281a6939c008f19082ee5d4d657b6b8b3028"} +{"id": "train_41229", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f711b09e61504412f9c32ce18950264ab67b8cd3"} +{"id": "train_41230", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4e9a8211cf4ee3fff9e0e8b884ad8c6fc6185f9"} +{"id": "train_41231", "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: 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", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "493d0d5dafb54380d9d408b2f1b731dff727e4fd"} +{"id": "train_41232", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0d17acae5b00b8d6eea8c5ce14e48b29d98dfe3"} +{"id": "train_41233", "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": ["repo_scale_reasoning", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54b5a7e829d80699d8e2b37667a8adb4aa17a36b"} +{"id": "train_41234", "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": ["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", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c66af4213edacb054705ec6bdb118cb0b66c2d0"} +{"id": "train_41235", "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: 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": ["governance", "auditability", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb65172e824691ebaf82f7696b5638687659cda1"} +{"id": "train_41236", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02ab5f7b977855eca6d4e48a65e372547ed06696"} +{"id": "train_41237", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ef659e84b0bf10a96c66422d8ba93ba78c05340"} +{"id": "train_41238", "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: 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", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb6f84bb3ee42a93e968a7c22d7f262c4960c4e8"} +{"id": "train_41239", "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: 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", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ee20d4ab15c0f0f77b3f9d924e969745891661d"} +{"id": "train_41240", "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: 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3635be21c13826c9d7e9758fa1d846a446418295"} +{"id": "train_41241", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fde16d633be6e66eb62077491b5e7450594e53f7"} +{"id": "train_41242", "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: 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": ["auditability", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2de728827fc60559fcc51d53138b928c467346ec"} +{"id": "train_41243", "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: 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": ["governance", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00e540845e92b4a1cd15a61bbf20cbc2ae9e7930"} +{"id": "train_41244", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01c6a988f94d03b97450b4059c38940ecd5872a3"} +{"id": "train_41245", "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: 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", "security_gates", "ci_integration", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "089a19431e2105f492a100522dd51fbfe7519f82"} +{"id": "train_41246", "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": ["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", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6a4420c3abaf22b63d2e5661d64dbf73c120728"} +{"id": "train_41247", "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", "cost_latency_tradeoffs", "tooling", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25e222563278a6732c7a1333c6186cbbb79beb91"} +{"id": "train_41248", "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": ["security_gates", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e14ecde58a422005bdc472d3545868337ba5e3d5"} +{"id": "train_41249", "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": ["reproducibility", "ci_integration", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7893cda23149f09da239786a9e61a02b71cb6c23"} +{"id": "train_41250", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["auditability", "tooling", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae07a5ebc590693fc860b82f569ed4a648022d75"} +{"id": "train_41251", "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: 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", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5b87c1382faf0a4c01f1a8a4a4342b22fb34bab"} +{"id": "train_41252", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f8cf9c2c9ed1c4a0db82f1d26271d45617ef685"} +{"id": "train_41253", "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": ["repo_scale_reasoning", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "beb41ec901300b8206653ab896d993c032b60c1d"} +{"id": "train_41254", "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: 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": ["reproducibility", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95e0aaa3953feacd4aa9ae4c65abdbcc8f1d8204"} +{"id": "train_41255", "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: 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", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e57304b7629c23f816918b082804b0f6c2d30e8"} +{"id": "train_41256", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10a2ab3c1037d5018482ee7bb013503b165bd487"} +{"id": "train_41257", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f0de66bfd4dca2d83d6552089b202932dfb65bf"} +{"id": "train_41258", "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": ["evaluation_metrics", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f058e1badcd4bbc176b9392029c6a4104ab9cc8f"} +{"id": "train_41259", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1742c868a30fd5090fcfa22b7ab97cc702cb7244"} +{"id": "train_41260", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a2b0e5deb1645b93165858a30f0a426cc4bef21"} +{"id": "train_41261", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "691b22fabbbaa2aa67389d162ff83e5ed979e805"} +{"id": "train_41262", "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: 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", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7aec9963ffee963c226c535df05f76ff7166c9c3"} +{"id": "train_41263", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "tests_are_truth", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee8b6c62166ea06da29f6551e19a61fb3150c3ea"} +{"id": "train_41264", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "security_gates", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25458ef71424d9a79fa2c1f1eac2519babba3623"} +{"id": "train_41265", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e678393b446d6979b1dce6f3a05f7051f5db3cb6"} +{"id": "train_41266", "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: 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": ["ci_integration", "reproducibility", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97b438857696ab63ba527bd6ea79692a9bdc8226"} +{"id": "train_41267", "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: 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", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a12baa89417920756c5c83bc38eb16cfd45efe94"} +{"id": "train_41268", "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": ["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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3543ea9e9b28cace873220a9d326c35e7d8711f2"} +{"id": "train_41269", "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: 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": ["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", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "febb7ef05b611dcd8a7d3a9ea535b4295254b402"} +{"id": "train_41270", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aaf223a5192fec848b33ecb2e04c26d1bc10eba9"} +{"id": "train_41271", "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": ["documentation", "tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "029879b4aad9e0bbc7df5f85971bdc48b7a71832"} +{"id": "train_41272", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c61dd0466eb131a47080a575cc4fcc3cfbbe6ebb"} +{"id": "train_41273", "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": ["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", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba53ef0be181bcd0437802f1b57107c4c04f6f6b"} +{"id": "train_41274", "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: 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", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6d2506a7c024aec78b3a3172bf8de5c98bb98ce"} +{"id": "train_41275", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "tooling", "reproducibility", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7e957bc76cf45f231fe4246e3d5622ebff12ed7"} +{"id": "train_41276", "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": ["documentation", "governance", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12cfa5859ea5797be010fd811b21d70dda3c962f"} +{"id": "train_41277", "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: 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": ["auditability", "tooling", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "188a73cc157053a40a3144a628729ba3e1219cfb"} +{"id": "train_41278", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8eab44793ef8cb2e6ce1936353a3b2b8131ea993"} +{"id": "train_41279", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["documentation", "governance", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b19cb36bf4519ebeb4afea455881524972b44c6"} +{"id": "train_41280", "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: 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", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb0b4559730179b19ac9a73f3c06a1ed77b41216"} +{"id": "train_41281", "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: 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", "documentation", "auditability", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec617e2ef94a9892917afbde829e64b534dd446c"} +{"id": "train_41282", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91e0b0fc203b5297366a0b8eea1728f2ff48b3fa"} +{"id": "train_41283", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d05bd8ad8bc8aa4e9512574d0d18dee4b265102"} +{"id": "train_41284", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01ba3baa9c6208551e0902d4a2d107cd5b43ea42"} +{"id": "train_41285", "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: 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", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee63f681409f67ea3ac605de89d644a14e865432"} +{"id": "train_41286", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb955e4850ba5fe17b792c7ea320577057302e4c"} +{"id": "train_41287", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1062ca2959dd0e793a24fd4eb8a43f392496262f"} +{"id": "train_41288", "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": ["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", "auditability", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "108f4e4abc1ba3116ef1c3f4e30c47f0a90646db"} +{"id": "train_41289", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "741b67a9cf16f70bfd54db1626626cd92f6b3787"} +{"id": "train_41290", "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: 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": ["auditability", "cost_latency_tradeoffs", "documentation", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7af68436a1be23fdc19e2f606849257e6d36394"} +{"id": "train_41291", "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: 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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0480ad0a4e03c294bf2cf23f49b2b922a8a77806"} +{"id": "train_41292", "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": ["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", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38e75bec6fceace823d7d0acf78cf8272d0384de"} +{"id": "train_41293", "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": ["reproducibility", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4e4c852428f688db3c8100c09cf100bfe4aea47"} +{"id": "train_41294", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fe8bd4c0b2fd3a82dc1f1a2d0bbc59b3e321222"} +{"id": "train_41295", "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: 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": ["governance", "tests_are_truth", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10f5b24fe4e43d5823abf3f9d7dd1eaadbceffcb"} +{"id": "train_41296", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "reproducibility", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "418c639c6ad717bbf0c1d191440c03fc8f473b1b"} +{"id": "train_41297", "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: 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": ["reproducibility", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67b7cd8c134f8c0629d37b0dbc3d321a3fb48912"} +{"id": "train_41298", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "405838758f38f1b8b4e6a81932a5514b316baa77"} +{"id": "train_41299", "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: 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": ["auditability", "ci_integration", "documentation", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4298976b520c9481abaadc11882fc4b944e7701"} +{"id": "train_41300", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["ci_integration", "cost_latency_tradeoffs", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99986bcb2b69b0ef148d9c71981b1b071317ffd2"} +{"id": "train_41301", "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": ["cost_latency_tradeoffs", "tests_are_truth", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b451d21804967448864db4551e30ce984fcd9070"} +{"id": "train_41302", "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": ["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", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cc18ccbcaeafd189d21b53ba750120c3bd34222"} +{"id": "train_41303", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf42c5c69d0e33b7884b3cd7de631753d99e995f"} +{"id": "train_41304", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ca3bb264c62bfb69b28a29debdace1a9603849f"} +{"id": "train_41305", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tests_are_truth", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7eec7c8baa3362c38dff58def4651dd1d5cc108a"} +{"id": "train_41306", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fab789501f1d0e2da04cc2b5c3b98784bdcf159"} +{"id": "train_41307", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd5e21b37e1f9a6c49d62caa0ef0a4a6ddad2191"} +{"id": "train_41308", "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: 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": ["evaluation_metrics", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a59c6c72eb00fa0b7f2b9b37c404668e41638ffe"} +{"id": "train_41309", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0d492b70ee432779394eb8f3be6c1cf64a15c60"} +{"id": "train_41310", "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": ["tests_are_truth", "governance", "reproducibility", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86e8feb58091ad990a926affed7fa3e9e6d3a5ff"} +{"id": "train_41311", "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": ["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", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6b33e3e336033e20bb2b61a35ba4a905382e4fb"} +{"id": "train_41312", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f32cdf3cd5581051665e4062fd1d41dc9aed2c6"} +{"id": "train_41313", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ab932f9869b4544c9210d909406357516208458"} +{"id": "train_41314", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b5f065f45088a243eca9c9ee8a81a002bcedc33"} +{"id": "train_41315", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e78e1fb025c6f0c6f59599d030b98b27c3ca826"} +{"id": "train_41316", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "auditability", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "451a7469450697924695b23a24c1b8ec9b582b63"} +{"id": "train_41317", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f52fd01813fcd34af4918fbf36295919e862360e"} +{"id": "train_41318", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5692af6f6c6730f456c93d0c7138bde0f924b074"} +{"id": "train_41319", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "documentation", "security_gates", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e2e8661445c6a6b5174a7ef028118e23a5805ad"} +{"id": "train_41320", "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": ["tests_are_truth", "auditability", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "326dbdbd6ca649c60d571ac1fae099a707172169"} +{"id": "train_41321", "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": ["governance", "tooling", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89580e238df2034ab23fb67305313727dcdcf58c"} +{"id": "train_41322", "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": ["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", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1aaf526d3bc9be5890181da419a62e54ad784d28"} +{"id": "train_41323", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "beb28f6a11a752c5b662b22608541077696cc425"} +{"id": "train_41324", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e95c2e374acad9a205f7df4152218f9d113944d4"} +{"id": "train_41325", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17bdbaa4d249777caddf394d5f28c0ca177d19bf"} +{"id": "train_41326", "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: 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", "security_gates", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8d0a06454bcb4aae42b9c6fb51af60297377fde"} +{"id": "train_41327", "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: 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": ["security_gates", "tooling", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fa5286bec93c5f9439cd743f9da70ca1b93b21d"} +{"id": "train_41328", "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: 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": ["ci_integration", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52c8079e530a1246eb4e1405804625110ff991d9"} +{"id": "train_41329", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "tooling", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e174230e1bfdfc8e4f0fa465ea21e8b77460d5fe"} +{"id": "train_41330", "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: 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", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d20188b4c72cb9b900967fe84b5c2cbd11fa83da"} +{"id": "train_41331", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3990fe06732a85c06c5ec65ad78d93d57b1e37ff"} +{"id": "train_41332", "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": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0da6a17b22abb23aa62976cfd09d734db4ca2b0"} +{"id": "train_41333", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready 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", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca2a92b77f3127df29728c7b9ed66a0c3e254efe"} +{"id": "train_41334", "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: 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": ["security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7411deb0a500b85515aae27ac5bb68b40fdeadcd"} +{"id": "train_41335", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["reproducibility", "governance", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "abe4f7067aef9813317234f1a04971b97a61b68d"} +{"id": "train_41336", "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", "security_gates", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddca09243334266497561e55fb424bceb4a5be64"} +{"id": "train_41337", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b26d08b331872df5b714737c4e73dea3c50cf5a3"} +{"id": "train_41338", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44713010d07ede19a148958d49af2a8f1aa9ec76"} +{"id": "train_41339", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ce037b1154fe5cfdc560a9fe94671ce9c007b56"} +{"id": "train_41340", "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": ["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", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d82b0cb525e5923f16ec323bd84830a515e6df9"} +{"id": "train_41341", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa38dafae1deb68de63804f69b7ed1aeb653fe95"} +{"id": "train_41342", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81dcb733adf387f68e716874643171970faebc27"} +{"id": "train_41343", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10ac6722609ed6b68a734f83c526f247063d4317"} +{"id": "train_41344", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc7379f5cd2f72fc52ac32c8a70de6b1811c8edf"} +{"id": "train_41345", "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": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e154c1dc7609f1eda62ea43220a8a56650c7008a"} +{"id": "train_41346", "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: 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": ["ci_integration", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ef4bcd09c82dfce99ee70f361422cccfbf3bbe8"} +{"id": "train_41347", "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: 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": ["ci_integration", "governance", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64f89c36267e78e28f6a566f41e30fb99a7df3ac"} +{"id": "train_41348", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07ac0d9bfe301201645730c275932514b498310b"} +{"id": "train_41349", "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: 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": ["reproducibility", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a30f657e60ed8c41d6928424a452341b92f2349"} +{"id": "train_41350", "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: 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", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcf3493090a56b7d11cc112227b5c4f14c595df9"} +{"id": "train_41351", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc1922fc2f958ab14617d704bed1ec7109953df2"} +{"id": "train_41352", "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: 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", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7068ceea9dc09516191bebcafc01e58fd46cfa6"} +{"id": "train_41353", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bb3586a2d8efab3ab6dc1c452696d440240a980"} +{"id": "train_41354", "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": ["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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "105a42a98ed2829cc300d9147d70d9ca653b8765"} +{"id": "train_41355", "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: 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": ["security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04879379981de9dfda0e199b92c238a080b3cde7"} +{"id": "train_41356", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1d1e10ef17e57ae88f7297b3cbcbf8cac7d35bb"} +{"id": "train_41357", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d67d9300656f36db54432fe4935d8cf8304b454"} +{"id": "train_41358", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ba2d9c02c05cd6c31158f947634a5788d9e5605"} +{"id": "train_41359", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75dfcb302745200ca0de4c0e56b915c6d3dbea3a"} +{"id": "train_41360", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca96d955bda05f14802a96be31ddb1242a749de6"} +{"id": "train_41361", "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": ["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", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2789583a0b0bd35b30afced641088fb2c70258b"} +{"id": "train_41362", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "governance", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff339ebc0592a4ceb74686ddae83154374ff2edf"} +{"id": "train_41363", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc4e238aa4b6b42f69c889265549077132858674"} +{"id": "train_41364", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "security_gates", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd439e093f587f53dace6ba5ed12e330ae92ad81"} +{"id": "train_41365", "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": ["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", "security_gates", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b239e94dc1c2bc462325b01b6a26166cc827c69"} +{"id": "train_41366", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12f634696dc08847aecada6aab4aea3d04e643eb"} +{"id": "train_41367", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "ci_integration", "auditability", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b90a38cd5aa10fa9d4af00339667f9d2c2965bae"} +{"id": "train_41368", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c93f129c7de0c3433f609f28744e2a1681f35ac5"} +{"id": "train_41369", "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: 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", "governance", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68464e14b443a36a39c9f5aa37d240af45f8102d"} +{"id": "train_41370", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e433d621b848c0138410b4ca86606e846e2aa095"} +{"id": "train_41371", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "tooling", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "516ff3cbd9d750aadc5b0dab149c3f935e67fc26"} +{"id": "train_41372", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e14419ef7660088af971de42e888e7bf6fe34eb"} +{"id": "train_41373", "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": ["documentation", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9eea7e74edbbfd109b09715c98bb566e6c686922"} +{"id": "train_41374", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26b382429baf4122af4e7a8867b1e50e82c8ecd6"} +{"id": "train_41375", "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: 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", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44cd33a06426f971bfaf580c57a932ce7b564a7f"} +{"id": "train_41376", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "712ac14ea61d4f5eeb3a03a64ebd6e911b8b3d56"} +{"id": "train_41377", "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": ["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", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69ec6417b6963909eeb29fbe1a21a02b7f4d39af"} +{"id": "train_41378", "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: 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", "reproducibility", "auditability", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b4a06ea43c357b616f7e3231671640ad6d004b4"} +{"id": "train_41379", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6349b401c4eed74fa0091947277346483ce80992"} +{"id": "train_41380", "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: 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", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f6e4dd51d4268f5c6c49e5a6669439c10574999"} +{"id": "train_41381", "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": ["ci_integration", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "143294b4747a05dcb794fddd70e34d27239cae8d"} +{"id": "train_41382", "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": ["auditability", "ci_integration", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed07d2af8b41133b143c0c53601b417bf3d54d96"} +{"id": "train_41383", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfcd5a5af2f128e755b8e54cffd801beff515d43"} +{"id": "train_41384", "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: 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": ["tooling", "ci_integration", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8ad3ab1f9bcc9412a94e4ef1738abd9a184d35c"} +{"id": "train_41385", "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: 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": ["documentation", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99a3dbd1dfc1bea93b0fa6c69f386762373f033f"} +{"id": "train_41386", "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": ["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", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa286d11b86730a34e3879439a8bd182a41d031f"} +{"id": "train_41387", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "security_gates", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3c4c2115b34bf3dab4dcf6fe9d2d2ca4812c2a9"} +{"id": "train_41388", "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: 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", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4dbc8ba41849c8c9a382db19bf78790360f5a24"} +{"id": "train_41389", "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: 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", "reproducibility", "ci_integration", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eadf88c1732253ba17f341a824dd9ac7509b01fd"} +{"id": "train_41390", "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: 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": ["tooling", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ebc897824efa75147612af6472b0b63e0bfefdb"} +{"id": "train_41391", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96ccfeb8a9fa1b6e75018d826c63baea8f0b49fb"} +{"id": "train_41392", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["tests_are_truth", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b687c44291abe1473d226f4784230e55323813fc"} +{"id": "train_41393", "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: 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", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1afc951546443690f56eb1ce6e57ac863d0e044d"} +{"id": "train_41394", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "tooling", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1681229f4834c232534e3fc5bd2bdee10b88a88"} +{"id": "train_41395", "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": ["documentation", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70765aa0c0f677a962b90deb950f0aea694a59d1"} +{"id": "train_41396", "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": ["tooling", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6d02d68514f5367b480b3226daf74c1f5c58ab7"} +{"id": "train_41397", "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": ["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", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e1b00da5daa5143cf9ed967fe3abee5bd4aa520"} +{"id": "train_41398", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d66e0a2f5097b32cdbae788ca1894434f6c3755"} +{"id": "train_41399", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fd211de501d936b3a431f682bb313041c0374c3"} +{"id": "train_41400", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4215d0c4d7741e911051c4aadeaeae91458357b4"} +{"id": "train_41401", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0617b7aae9d3c963a6ff255efb7ccbe7304ac2a0"} +{"id": "train_41402", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a45dc7a6dab05cb3057257e781cb449e84cfb55"} +{"id": "train_41403", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6db75d87c14af95ea5b747d60e00096e5bee67cb"} +{"id": "train_41404", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c9fc0bf9549ce1473b034bc1ec6eec728edb2a5"} +{"id": "train_41405", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba60b023c149a070cda1bac0477c76296267fa8a"} +{"id": "train_41406", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d9c3f7fc0d6d84748d58a98f0c07ec14c29d47b"} +{"id": "train_41407", "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: 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": ["ci_integration", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b10951bfb8e8237aed7f8a1f2c417447375d298"} +{"id": "train_41408", "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: 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": ["governance", "tests_are_truth", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8f5c4fe5fa9aeb26505ea433c52eff3d00d0d23"} +{"id": "train_41409", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c03e99acb2b9e8afe3af2c4e886454df655b2d43"} +{"id": "train_41410", "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": ["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", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59d974c5117a9486a7345296f7759fc19a6c32ed"} +{"id": "train_41411", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96629ab613117190631bf8ba25cab1cb21873a9d"} +{"id": "train_41412", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a75e610ddb2ccc87b67c4bd8199e34db752def81"} +{"id": "train_41413", "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": ["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", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3aa2069eb178b66d0c560e72fdbeb274edd7c17"} +{"id": "train_41414", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e771df439b2303a0c757eddc7dfd944bf599eb11"} +{"id": "train_41415", "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": ["security_gates", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23da9fb2a25140cbf4b1e9650f857a8aef865eba"} +{"id": "train_41416", "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: 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": ["tests_are_truth", "evaluation_metrics", "documentation", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ea8aef99d5e6e5a2e3f6fcb587b2a8205574c57"} +{"id": "train_41417", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62dbbbd87323ab85b4fd7528d6dc5fd514c8babe"} +{"id": "train_41418", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d64ded8315be8d18059a8439daf049670c61fee0"} +{"id": "train_41419", "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": ["auditability", "tests_are_truth", "security_gates", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd298fd15ba61c7dc8b967a256c9f4c31305c638"} +{"id": "train_41420", "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: 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": ["security_gates", "evaluation_metrics", "governance", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2dd0fa5de465a1d54bd2e93a3892adde24c0a9b9"} +{"id": "train_41421", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50c9d59fe5a67e7b35317113d76ad5e6e0043c30"} +{"id": "train_41422", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c87e33089a9934c6012ce850b3deda09e9561e61"} +{"id": "train_41423", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d1b0f464d8d520dcaabefc6a4231e8f732a1e3d"} +{"id": "train_41424", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eae855e82252b0268fedf0b80b0a4f3e70dae5f2"} +{"id": "train_41425", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16f642c4804837577fc607d942cf0428fdf85643"} +{"id": "train_41426", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92fee0ba901177d655039a5c70c6fc9c024a5937"} +{"id": "train_41427", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8e12f098edf95ad8140b0f9afec75855b624ee7"} +{"id": "train_41428", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "repo_scale_reasoning", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94d82f3228f02146a38a974767b2bc57108361ab"} +{"id": "train_41429", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d86d6979aecb8931222d7623d956f461c45358f"} +{"id": "train_41430", "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: 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": ["documentation", "security_gates", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3759674716b4bef1c5d64e7e427b521852bcd38"} +{"id": "train_41431", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9bc93c7ffb9eb2eec10a08a35e688e25931a9b14"} +{"id": "train_41432", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca4bd3b68d1840796c4efffe4524d2e0249b0c3d"} +{"id": "train_41433", "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": ["security_gates", "repo_scale_reasoning", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1415d596d5c74eb88b319035fa571a6312f20b13"} +{"id": "train_41434", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f8961f208b2ef6663ed95e3b6c9ecd3dff665c2"} +{"id": "train_41435", "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: 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": ["ci_integration", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3127292c51a712a3e4c35da06bf89470e58992a4"} +{"id": "train_41436", "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: 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": ["tooling", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69b9814858e16a42242fc163c2ae4e838ca81fb5"} +{"id": "train_41437", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1971cc4ee490740e637f9d65f2e0af2d9ecd006"} +{"id": "train_41438", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "062b0af04a23cf4092ee0ce311e8ff27ff4ad3fe"} +{"id": "train_41439", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a95b967bd2368d72f2ed2480d55afc2bc2b27a30"} +{"id": "train_41440", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["documentation", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "929446e2c15f1ad2acc65c858e7bde748b4c911e"} +{"id": "train_41441", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f46441a94792cb47d69fc6c44a760559da5a552d"} +{"id": "train_41442", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a86da61187b5c0caab8701754a357ca827a782cc"} +{"id": "train_41443", "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: 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": ["auditability", "tooling", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8ea2754c64d0e94c5d044120d3222c0b48e2724"} +{"id": "train_41444", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f09c467164b8ed5891abba655053cbc790147d9f"} +{"id": "train_41445", "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": ["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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63cbad3cab0401039293db79fc41b99502566483"} +{"id": "train_41446", "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: 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": ["documentation", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28229fd2bc9f6355529f066203d774e2987706db"} +{"id": "train_41447", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "259d8cc7fb07aa8767c878f0e34cb15b6f009154"} +{"id": "train_41448", "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": ["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", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "823e35fa0431c38a2c8d4b867b528edca25490f7"} +{"id": "train_41449", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd39546701d6ae6f004373810ee99102b9c67be1"} +{"id": "train_41450", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44174c01baf798071a5ea1d6a3a5590b69a4aa12"} +{"id": "train_41451", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40309b458e48fd8fccdf045fca3daffae7e2adca"} +{"id": "train_41452", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "077762be642420f7683c458f68bc884d3ce084e1"} +{"id": "train_41453", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "documentation", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19740bb4fcda293338a310b073915b45fe9a86c1"} +{"id": "train_41454", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cffe2e945e0b3e183a02775fac562294e873806c"} +{"id": "train_41455", "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": ["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", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2cff725c5c522b7abb6f9b42c044083c94faf55"} +{"id": "train_41456", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "governance", "ci_integration", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c82933ed06c129fec2e04b874f85d74c35fe33e1"} +{"id": "train_41457", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7046562a89f3ccf949aae9d74035aef0f435870b"} +{"id": "train_41458", "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": ["reproducibility", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e44e8e5280e067d801d7126035271cad15d0afbf"} +{"id": "train_41459", "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": ["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", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f148c5f1e660df010309f5cf065ed3a0de610877"} +{"id": "train_41460", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c51d0fbe3d3cbf5e33148c3b7633c610be1e0bbb"} +{"id": "train_41461", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "security_gates", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae556a09e86d9170118acc22dd349363428fc783"} +{"id": "train_41462", "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: 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": ["evaluation_metrics", "ci_integration", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa6f3ce8db353147f1794a2e0fa71a99ffc1d5c8"} +{"id": "train_41463", "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: 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": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "515ef065cc1184a01589af6c4741686330d6f97c"} +{"id": "train_41464", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6011afc3016aeb721a0d3d3a9cc4921ebbb45a03"} +{"id": "train_41465", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43fcfd556d5e00182c71b2736a81f26e875f0681"} +{"id": "train_41466", "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: 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": ["reproducibility", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75fccde20fc7b45756c0aebd6b85240d9279ed17"} +{"id": "train_41467", "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": ["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", "tooling", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5752fbcd3793911cf195d52ac1ef91756fac9149"} +{"id": "train_41468", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["tests_are_truth", "auditability", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cfb03267e29864b417b5c703211d6764defcd87"} +{"id": "train_41469", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3e78693f1280e25526350aaa742ba386e258583"} +{"id": "train_41470", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffb7a67a41e84e1d6e3f49b66fe61418b56f3aaf"} +{"id": "train_41471", "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": ["ci_integration", "reproducibility", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a490765279bb9f6a4f9d63e5d28b021052cad560"} +{"id": "train_41472", "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": ["reproducibility", "auditability", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08ccc37c9abad4ba5dff485e28d1f0e7d4b37716"} +{"id": "train_41473", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "feacf0f3d59768a8421e1ac22fcf9a9e8fb53fd4"} +{"id": "train_41474", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "468a738800c47e1c957c02db2a8b50918fdfdb4d"} +{"id": "train_41475", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8ae01fdf451ac8c733813dcbd59429c8b1b0389"} +{"id": "train_41476", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84a38e78494a3a6b95ddc9a828305ae6286a202c"} +{"id": "train_41477", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51f2fea29357f3bf2f7d0a9d55cde8775f0a5d99"} +{"id": "train_41478", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9650bf9e318968463320a78eac5afe00dfcb6b60"} +{"id": "train_41479", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c11127ea99ba814de6cc48d62197c0eb15e06e94"} +{"id": "train_41480", "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: 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", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "adf25af28747d5e524d119ea3876201befa1ec8d"} +{"id": "train_41481", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4dab624fa552ebbf1115d5781055806be174263b"} +{"id": "train_41482", "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": ["cost_latency_tradeoffs", "tooling", "documentation", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7cfef243890c2de26e3cf07b446ef32ebe8a0c1"} +{"id": "train_41483", "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": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2971397baac89d3d00cbc258453f08cee33f7afd"} +{"id": "train_41484", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa13909170668c2cee69969f015581ebc1ea9333"} +{"id": "train_41485", "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: 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": ["ci_integration", "tests_are_truth", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3902d32f6becdf32cd25b0f373f7ceffc4754372"} +{"id": "train_41486", "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": ["repo_scale_reasoning", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34c5122b6ea3ebf4d89d152ec30bb3b0ba1b3602"} +{"id": "train_41487", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b29853a339741678a579dd988f9d20c5a5f46a59"} +{"id": "train_41488", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "documentation", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f13431bd7baca643d3dfd392700ff1462361cfca"} +{"id": "train_41489", "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: 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", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c564e9527906ba4da4d10d479792ea7c4189231"} +{"id": "train_41490", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a56a5b82151bf4ae86a4c5aa2282a7f8e18b1120"} +{"id": "train_41491", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b01d2f4cc79e2133c3007bc5fc8277fcea48750"} +{"id": "train_41492", "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: 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", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efef7672cf8a6498ec161ab57ff0c79b8ed87f09"} +{"id": "train_41493", "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: 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", "auditability", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb10dd3d892dcfae7c2505d6f8f2e7011cae3016"} +{"id": "train_41494", "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": ["repo_scale_reasoning", "auditability", "documentation", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2750524268a79692ffe21dd8880276942ac8d15f"} +{"id": "train_41495", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19740bb4fcda293338a310b073915b45fe9a86c1"} +{"id": "train_41496", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["tooling", "governance", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38ce036abccfed75fe3abfcd8d07f032e5714d49"} +{"id": "train_41497", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98fc2fef1e6075c48b7e6a0708797dbbc668ab0c"} +{"id": "train_41498", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92788372a172f6ed02628a5cdf589cf477046908"} +{"id": "train_41499", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "tests_are_truth", "governance", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "863344129d739dab97e78a8fc2615c7cd8f5149d"} +{"id": "train_41500", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6dc37432139cb39cff0aa6509da32c2512616b1"} +{"id": "train_41501", "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": ["repo_scale_reasoning", "tests_are_truth", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31b70315c7753574af5243747a0196d0182d4470"} +{"id": "train_41502", "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: 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": ["documentation", "reproducibility", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d76e3f1ac7130f736644d3167e4480e464ae39f7"} +{"id": "train_41503", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f3f38110ba0c93d0c87c4077992dea8ea847db3"} +{"id": "train_41504", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50d9d6b73fc8734a2fc93f68c85363fd65ed8fab"} +{"id": "train_41505", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d4f078496234fc9fc4b58a491b705c623e66e9c"} +{"id": "train_41506", "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: 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": ["reproducibility", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80d4574d5555eec06d07e0a9c9ce0dedbbb8c1cb"} +{"id": "train_41507", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec5294ca12bf0af24429a0d3d79aff45d0494923"} +{"id": "train_41508", "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": ["security_gates", "evaluation_metrics", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b3b39d29d50323687f31cdd02ba47dc558580b9"} +{"id": "train_41509", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1ffbcfccd994f02dcddb55ea478fa91bf310587"} +{"id": "train_41510", "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": ["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", "governance", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a531ddb5932470d5925dc6cbdb809fb3b208199"} +{"id": "train_41511", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d8ead2c234b5cdcde3512e42cdf0e93b9c32cb5"} +{"id": "train_41512", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99347eb5719ed3678b5ba6ad5eb0fc72a788229b"} +{"id": "train_41513", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62a583cd38b9ec94dd8c30f52c386a27578d2d2f"} +{"id": "train_41514", "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": ["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", "reproducibility", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a4a5048ca10e7709d6df5853b33ca894d2130cc"} +{"id": "train_41515", "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: 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": ["security_gates", "ci_integration", "governance", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4f06195267de0e693aa1b37df2f4c8f771b0c4e"} +{"id": "train_41516", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ab99d6e8aee23ca792b710ae0c1e6aade107724"} +{"id": "train_41517", "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: 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": ["tests_are_truth", "auditability", "tooling", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "863ff1b0deec664405e3806b051eff71521388ca"} +{"id": "train_41518", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4379adf174aaa657dc98fcf734d986dccdef72da"} +{"id": "train_41519", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2cbc2a69a373fca0387a2cab3e4bc89e140a9d1"} +{"id": "train_41520", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b0152fc867cb73ce7f0a370c4069462bcf7cbc2"} +{"id": "train_41521", "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: 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": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38a9ffea607cb3c280d66be7452dd5208cd44e6b"} +{"id": "train_41522", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["reproducibility", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a04a7f98a48041058c832c15d72428f43f88ae9c"} +{"id": "train_41523", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "838847b678bb9a097cb4ad91188e85d54e4f730b"} +{"id": "train_41524", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "auditability", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da803d32c852c30efac8f73783c446a5e0ef409a"} +{"id": "train_41525", "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": ["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", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4dc2ff91481d76d165df2956108f88bf3bbad8f7"} +{"id": "train_41526", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "056936413eb9083ed2df7f65eff0cd5be0b25a83"} +{"id": "train_41527", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b58cd45a1eee01bbc8a02d8f46094bb59d6ec3d4"} +{"id": "train_41528", "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: 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": ["tests_are_truth", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e881a405a0217a7857e191e4894622ce1ee6073"} +{"id": "train_41529", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0008fabc5a6c94e38caf77c3fbbfe5fa13768973"} +{"id": "train_41530", "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: 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", "cost_latency_tradeoffs", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73b19ce1ffce8c3614733c81aff21642a174dc01"} +{"id": "train_41531", "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: 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", "documentation", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94439c961881e439c441ebc404d771802c00e760"} +{"id": "train_41532", "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: 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e17ae1c9bc7116b3d77dbdab3e994aa7ea939d79"} +{"id": "train_41533", "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: 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", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08683bcffa9c3403019a09918c60dfc920577b45"} +{"id": "train_41534", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c556ede5704a3468639f4a6023b0a3dfb2544fb"} +{"id": "train_41535", "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: 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", "security_gates", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de8b87756832be47d89ca31867bfe4bc7130b44f"} +{"id": "train_41536", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd092b753e1edf209bf24f0828d317ff522d7cd5"} +{"id": "train_41537", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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": ["cost_latency_tradeoffs", "governance", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2244b7fc5aed68b657d1ca6609ae268224497054"} +{"id": "train_41538", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1b3716a1357a90fbb88f24a811bb16ecd76d07c"} +{"id": "train_41539", "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": ["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", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "512b8bd43a126e6758ac4c47fa6e67bc70f25eec"} +{"id": "train_41540", "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: 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": ["ci_integration", "documentation", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9d1f12351022ee3430648f7497f3a90757c3c78"} +{"id": "train_41541", "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": ["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", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4a3c88187e7ddd2dc9c6c7d2dd2bf901b1ed39b"} +{"id": "train_41542", "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: 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": ["security_gates", "governance", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8c346b5bbd029a4d526e0b5d5150bc88b637a1b"} +{"id": "train_41543", "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": ["tests_are_truth", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04f7b268cca04654060580bbc1ed6dd1416ec10b"} +{"id": "train_41544", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b38b3b6cdbe03e8e61264579e70b88d08d02831"} +{"id": "train_41545", "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: 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", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a14a3086f259b1680cd23ec701bec26d1c36d74c"} +{"id": "train_41546", "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: 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": ["cost_latency_tradeoffs", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88e3ac67cfd1ff77097905a53a8160ae379d9bd2"} +{"id": "train_41547", "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": ["evaluation_metrics", "documentation", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f40db5a87779abe48be3c4a4676c91482da32f90"} +{"id": "train_41548", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc558c485b17f3d29045302ca0d5ebd254598b0f"} +{"id": "train_41549", "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: 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c34e1a6e07a2d3889323af8914f0a3d6df78ae3a"} +{"id": "train_41550", "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: 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", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47470b708c1f169a6f55a5b91cd3beac7f6488c0"} +{"id": "train_41551", "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": ["reproducibility", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8bf0c7e33e75f50f5f53ca58357d1e7141cf898"} +{"id": "train_41552", "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: 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", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "594e94c7cb3ead7f93d01bc504037e0d6aec6d82"} +{"id": "train_41553", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d15d8a9a2ff116bcc21e082f3a7e6beb7035c655"} +{"id": "train_41554", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d69c69e1da7edee86d64e6738b4f853ff9be0355"} +{"id": "train_41555", "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: 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": ["documentation", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d0b9077c3fd72656327bab0f8dd2c7f45a9c9ea"} +{"id": "train_41556", "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: 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": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a568f36437004accdafea557d5ac63698cb836ea"} +{"id": "train_41557", "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": ["tests_are_truth", "ci_integration", "auditability", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff2c967f8def0c9f196dbc1073a1ccf7f249ca0c"} +{"id": "train_41558", "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: 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": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d2bdab8259e529aa28276704601e47f9a31c65c"} +{"id": "train_41559", "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: 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", "security_gates", "governance", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6ec48c4c10b8be47d6165e45830487e607dce36"} +{"id": "train_41560", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "372a594efe0dc214730492ffe4a5caf6828c19d3"} +{"id": "train_41561", "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: 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": ["auditability", "governance", "documentation", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7ba571297cc2c6a5eff1191042c16d14d116d70"} +{"id": "train_41562", "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: 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", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22d8023af9ff7ad1656ec76a6132525aaa286534"} +{"id": "train_41563", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9673243d155f98246d1e91e3cbdd2580a4542b1"} +{"id": "train_41564", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3364ee5b15c2ce6160c1044cdbe94c5d9dce62c"} +{"id": "train_41565", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c78ed3e7b3a03f7b82098c69b6133764c4068f68"} +{"id": "train_41566", "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: 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", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aab50aa8c1634ddcfc47aa69591c970b142b137c"} +{"id": "train_41567", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98afd21e4da187e5eba8ecf280279a403274d38a"} +{"id": "train_41568", "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": ["governance", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a63cdd152c163f5905a8258f3889e176889332a"} +{"id": "train_41569", "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: 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": ["documentation", "tests_are_truth", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23fffffc542533e3a6dab0c103a5d2571c2fd05b"} +{"id": "train_41570", "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": ["security_gates", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a881723bdc47aa242a457e8178417968829525c"} +{"id": "train_41571", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e2a7ea1afa4d8c98912d4ad1d1412683fb046c0"} +{"id": "train_41572", "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: 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", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c6c0b0fe8e0e5d78e2c133414c6a528a4a836e0"} +{"id": "train_41573", "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: 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", "evaluation_metrics", "security_gates", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "790527641a9564b932c048e9e01ecb03044f4bbc"} +{"id": "train_41574", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fadc3664022d793f1eadeeb576780fdc1c420948"} +{"id": "train_41575", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "cost_latency_tradeoffs", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac3116deb6e86abee1d50218eb0c3cf629d4ebf4"} +{"id": "train_41576", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["security_gates", "ci_integration", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c490b159f2183c00e9f987c8fe3f2604887f419"} +{"id": "train_41577", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "864310f43972cc6ffaa280f132e2844b9c045495"} +{"id": "train_41578", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "security_gates", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e34003a28d0f2eab6d7875e1e142b04849bbfee6"} +{"id": "train_41579", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eaf6e755064f20d4dc09848a149eb28ad28107b5"} +{"id": "train_41580", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "ci_integration", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e678393b446d6979b1dce6f3a05f7051f5db3cb6"} +{"id": "train_41581", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2596f136332ab9e4198ce64b24020a62f0ae0e1b"} +{"id": "train_41582", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09e180990e810438b6d8334559c7d5af2e5ee754"} +{"id": "train_41583", "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: 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": ["cost_latency_tradeoffs", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "215b0a8e0393adbbda41529b4f09ce711862686a"} +{"id": "train_41584", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c37f55bc85f289155c534c080c71b501c503e91"} +{"id": "train_41585", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8c1bfbdaac774e3fb8555297ddcf4c41b4ee92a"} +{"id": "train_41586", "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: 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": ["reproducibility", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23b1cde87f904d69228457f797401a2f41682302"} +{"id": "train_41587", "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: 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": ["reproducibility", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6145902264e1cb78066bf27e0b1a5ed7242be178"} +{"id": "train_41588", "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": ["tooling", "ci_integration", "documentation", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c072e4dab116010a007e1e167f62f3685d0649e5"} +{"id": "train_41589", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "476fbb7344bd3f8340746c6623eeeaba176fc9c4"} +{"id": "train_41590", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f85ab7e42b34059b912dc73abf93ee798022090"} +{"id": "train_41591", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f88163b1b1896ae680c25d6b88d185697d6e045"} +{"id": "train_41592", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa7259912dd82d7114cf421d0c6c296e040f51b3"} +{"id": "train_41593", "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: 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": ["security_gates", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26d0c1fb237345bf49fd358c40541f602b3ae7a7"} +{"id": "train_41594", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e4698825efc44227769e2b7f2b24185b58d04fd"} +{"id": "train_41595", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "tooling", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7fc4e3507842ea946fc8f763cc31aaee0fa5c88"} +{"id": "train_41596", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "governance", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f6f69f13c30168adcc9740a41c1fe69250fad3b"} +{"id": "train_41597", "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": ["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", "governance", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0748da530f25c1bd70021467f74f19070e520e17"} +{"id": "train_41598", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cc2294bb74bb1bc444f2e15288ac92bd9c26a70"} +{"id": "train_41599", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c678188630228e2a0096570432ff8a0519012ec4"} +{"id": "train_41600", "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: 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": ["auditability", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2334c9a28f49c99d10706d29a2859334eb172746"} +{"id": "train_41601", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "security_gates", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfe2d3148d1b2362ef4db4a067123fd347c9d920"} +{"id": "train_41602", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d5d3142021d4c8ac0af0b061c73feab2604f933"} +{"id": "train_41603", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["auditability", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "225bf45495973ff22f198549d6b06ec7bd201f51"} +{"id": "train_41604", "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", "reproducibility", "tooling", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35ef3a442275611647dd4ad916f85879995fde9b"} +{"id": "train_41605", "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": ["reproducibility", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49f8c41be431b68ab1b163316cfe4896ef844f00"} +{"id": "train_41606", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12f14dbb7c5d5646e20e6c2e98095fe19c287ef3"} +{"id": "train_41607", "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: 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": ["tooling", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6290f9624b24743f2031e406a461f4c306cd44e5"} +{"id": "train_41608", "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: 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": ["security_gates", "repo_scale_reasoning", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37bff01c9311cbe3788fdf08bdde43f77f79c366"} +{"id": "train_41609", "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": ["governance", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f60736714a81662d163f2b7d6c313fa1bada285b"} +{"id": "train_41610", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "168557b7a25794c4fdfb8b0ef2455230c365f389"} +{"id": "train_41611", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "352fc7623662d2f5d300746eaead465d9a92572d"} +{"id": "train_41612", "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: 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", "tooling", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b747bb149ec927b41d064b944d54cdde61cbeaa"} +{"id": "train_41613", "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": ["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", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d57b5392c1611bc37b5673825f1c37ab4e54922"} +{"id": "train_41614", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "documentation", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bf1d0513d4d885d1ed5f7e05a214f59b99efe02"} +{"id": "train_41615", "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": ["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", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bbeead378deeb68bc7282336fff0d0b011907fe"} +{"id": "train_41616", "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": ["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", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0b12d2e86d43629ee236be116f16f0afe34aa4a"} +{"id": "train_41617", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "governance", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d31f8b1e52db87c0ef408a68c4182013e63b71f"} +{"id": "train_41618", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d76fe65abb0a37c4b0d73bc7e6d2b598054c4138"} +{"id": "train_41619", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89867d91222280f37152e2d1fe5fdab704bbf794"} +{"id": "train_41620", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["security_gates", "ci_integration", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05ccd41fd00bd56bcf22f137000d76e4c7b13162"} +{"id": "train_41621", "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": ["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", "reproducibility", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f921a2dea01cef105423d25497b39436b38eb1d2"} +{"id": "train_41622", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08075d1769c9b7f650509805a744455b38092779"} +{"id": "train_41623", "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: 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": ["documentation", "auditability", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fa409aa48f41599130a251565ae034fba93f6f7"} +{"id": "train_41624", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f8cd9e0292e19e262aefdcbb37a90bdc0a96670"} +{"id": "train_41625", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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": ["security_gates", "tooling", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "866254a7dac49c99f9fe7da7d7b66ff5666f84e3"} +{"id": "train_41626", "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: 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", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddda3a31cac6ce880112ac4b8b8c04a3ee5f9ff2"} +{"id": "train_41627", "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": ["security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96248cf5e9d92f7d41d6121d743a5b1f19e0cb3c"} +{"id": "train_41628", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e01f71883eb85bf621f33bb5c5a86fcc8669c7e"} +{"id": "train_41629", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c28479b5f99e1547588c7c723fe9d2ad75b51408"} +{"id": "train_41630", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "documentation", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01d91e640ac56e11230192a1cab8edee840dbd40"} +{"id": "train_41631", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "486b9a395b67fb6360e3c1156d0fba9777643d8b"} +{"id": "train_41632", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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": ["evaluation_metrics", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8384f1a2ba44310ff825f58ad18ee6e731cd355e"} +{"id": "train_41633", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["security_gates", "tooling", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ca3bb264c62bfb69b28a29debdace1a9603849f"} +{"id": "train_41634", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bfdd87d34562e801cc3c4713b8625d80ab073d9"} +{"id": "train_41635", "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: 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": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3c1630f0dbc80c77024eaf902095ce13c3e09ae"} +{"id": "train_41636", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1062ca2959dd0e793a24fd4eb8a43f392496262f"} +{"id": "train_41637", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["reproducibility", "tests_are_truth", "documentation", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9816a15f05ac1c78be8e67a4e1cece98850a3bf8"} +{"id": "train_41638", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73839233eafcbf43351472b537f0d9981ffa02c9"} +{"id": "train_41639", "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: 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": ["evaluation_metrics", "security_gates", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0b17fcf1940e02698491749fe0a524f4d05634a"} +{"id": "train_41640", "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": ["security_gates", "auditability", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c856b68378e0630e1c2953ea616c9e0732c383a3"} +{"id": "train_41641", "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: 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": ["governance", "tooling", "security_gates", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c564e9527906ba4da4d10d479792ea7c4189231"} +{"id": "train_41642", "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: 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": ["tests_are_truth", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2902dd62d96ea94d51278664a31c4a67d5a3354"} +{"id": "train_41643", "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": ["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", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd377559eecb6115d114fdb8bacb554f214e7b3e"} +{"id": "train_41644", "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": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6bc9ed12ff71c364809defe21a623a16abef67c9"} +{"id": "train_41645", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dafda0f5e7844ac27f64e732cf97308ea99e51d5"} +{"id": "train_41646", "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": ["tests_are_truth", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "adf12eade4082f92ca4e81a06f6402c73487ee3b"} +{"id": "train_41647", "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: 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", "repo_scale_reasoning", "ci_integration", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7cada3c6d4157252f15875e7b88135b605fc3a1"} +{"id": "train_41648", "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: 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", "tooling", "governance", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0dc6791df8bbda4356b33bb41af14b439e93fd40"} +{"id": "train_41649", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "531771e97a39e0288dd7e1e0c8875f79297e76ec"} +{"id": "train_41650", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "367fd52289a15748a68acc7ca439c3c19bbb6b89"} +{"id": "train_41651", "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": ["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", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a6a931fffd7071107bf73e33b92e4e73d9df778"} +{"id": "train_41652", "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": ["tooling", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09bd3afff29dd4fd083a13a211e23a0db71974b3"} +{"id": "train_41653", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "932a63bf8fcb16b1dd6163fd3e2a2f5241c3ba7c"} +{"id": "train_41654", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef9d782e1487984774d5abf4a6579dbcfef14fc0"} +{"id": "train_41655", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "728d24ba9f9fbc80401333d1446cd92fcc1fd414"} +{"id": "train_41656", "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: 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", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a59d1485e43b705ed1cb3ff1660562efdecf2f45"} +{"id": "train_41657", "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": ["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", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3758388bf2aa41c107006c3837622bea24ba45a"} +{"id": "train_41658", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d467e0e9226df253aa4c1083a3f2f3e6dc333e40"} +{"id": "train_41659", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73ebab5ac46cafac2c10fbb9288f2bd56b7f6418"} +{"id": "train_41660", "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": ["evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9e4e0faf7d4ec8c821f2febfe3d100e16af3b3b"} +{"id": "train_41661", "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: 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", "repo_scale_reasoning", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bea60fd5d007e583b43400c4afbdb27f4ff8d86"} +{"id": "train_41662", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "935ec4af441f89bc7a09380724cb617490a2f10e"} +{"id": "train_41663", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "tests_are_truth", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "847bdf9f0b3fbfab58eb88ffc4069b56924fd4d2"} +{"id": "train_41664", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb38269707ae83173515a4dfe190fb0477883f49"} +{"id": "train_41665", "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: 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": ["tooling", "repo_scale_reasoning", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daed6a87b58cf573d556359cb2c0f5aa06ac335b"} +{"id": "train_41666", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tooling", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18a90d01e0d232e17f92ac6a812aa9c66e7dee17"} +{"id": "train_41667", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c26426a8d2c2dada26dea9f92325cd89f7aa86cf"} +{"id": "train_41668", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14ebfd50b60d314517399ec8fd9b48fee719a235"} +{"id": "train_41669", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "297bb6ce10d2b47b155a9974d532e0d5213648c5"} +{"id": "train_41670", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b27429eb830655075298574c61365a991ee1064"} +{"id": "train_41671", "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": ["documentation", "cost_latency_tradeoffs", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eed67ba6cb3af11e05409f18b290cc43f95c74be"} +{"id": "train_41672", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "reproducibility", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44e21cdad21ea2313df2078dacf18e1eef9c7214"} +{"id": "train_41673", "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: 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": ["repo_scale_reasoning", "governance", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23424c81e0d8c4da2c5d6f0c3c0be2f9fc39590e"} +{"id": "train_41674", "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", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2d66175c49d98f2491e806b365fd77ff0662b24"} +{"id": "train_41675", "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": ["security_gates", "tooling", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23fd7042780392ebbc1584ab50a5ea1d5bb1c545"} +{"id": "train_41676", "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: 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", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0238a66759834bdd64c39e1bb8fc1ce132a8c087"} +{"id": "train_41677", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tooling", "governance", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cb13ce8d15420f9eb018b0f4a5c70ade7a8b07d"} +{"id": "train_41678", "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: 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": ["auditability", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "408c7522f750a81cf134b4c3c8a66d94b0307282"} +{"id": "train_41679", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "080982ea5b1ebc0ff7c54c33d881461131c7d6b4"} +{"id": "train_41680", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56a5bf806b7346b79f6c46fd190c9267fdf5e103"} +{"id": "train_41681", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c43dc918f9ee175f405c246309eb58c1cedc2888"} +{"id": "train_41682", "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: 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", "ci_integration", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "298ada66a996e821fe13f0b85b2a2067b4652753"} +{"id": "train_41683", "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: 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", "ci_integration", "auditability", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "008ddd2d2762d6477f159a2a97d21b571ac32c8f"} +{"id": "train_41684", "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: 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": ["auditability", "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "213294603830d300af1c583726f4d323bd38229e"} +{"id": "train_41685", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba76bb4bb15a2d25699a90a2c41338a238fddd7a"} +{"id": "train_41686", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a76ce1daa46494db5f97be814f0c626fbc7c198"} +{"id": "train_41687", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bd3bc8c76e151f1514109a459dc2e3db347ec8a"} +{"id": "train_41688", "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: 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", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dda566bb201d11fc1a2e79202650f461fcd75764"} +{"id": "train_41689", "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: 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", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28e5b6cf52e88ee4e83bdd567b11fe533402645d"} +{"id": "train_41690", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a339fd28ce799b38b149aa2c32565fc2aa7d378"} +{"id": "train_41691", "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: 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": ["reproducibility", "security_gates", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e937da406824750b4370ddf44235d948f165175"} +{"id": "train_41692", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fd0c78d9ec0d4546c1c2de6568df2f6ac66452c"} +{"id": "train_41693", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd35235f9f465bdff5c1dc74fce66532b7251e1e"} +{"id": "train_41694", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ea2ad84d4478541b65317309c6a29edd1400ee8"} +{"id": "train_41695", "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: 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", "documentation", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85e1c393e4c6a23397d06bf5d4f30bd3e8fa09a3"} +{"id": "train_41696", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0bb1cb3b4d3af417ab7d74585ce04dc1b0f38d19"} +{"id": "train_41697", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc45d795a644c7a958299d66acb2d5804540f0b0"} +{"id": "train_41698", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b5e3b55847e53131b5ec42609d4f90dac7fa628"} +{"id": "train_41699", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "documentation", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34f94aa8208c262587f82aaaafe8e120baf6911c"} +{"id": "train_41700", "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: 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", "auditability", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18b769580000862c41551c91bde652c05fec70cd"} +{"id": "train_41701", "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": ["governance", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36a54b4db525c8afcd2ff5181845040f072c52ca"} +{"id": "train_41702", "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: 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", "governance", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a517347d949d018c0dd46c9d9990e4c2d30c4e91"} +{"id": "train_41703", "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: 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", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5fe14db5d9bcf93b71251a1cecec5c6b8689293"} +{"id": "train_41704", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a8c9019f3205fbc26ad29f47e2c3dfe920a3ac1"} +{"id": "train_41705", "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": ["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", "auditability", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "737ff1ad97255a9cecf2fe17ffe1b77c9ea5568f"} +{"id": "train_41706", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44ddc36bfd1538c45377b831e5ba89745e6f0122"} +{"id": "train_41707", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c14e01c1033f52ef67d1f9bb3ad2b84a6c471616"} +{"id": "train_41708", "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": ["tests_are_truth", "documentation", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8444e6e6e9d916333b9ceb22fdcbbe8d8cb90e94"} +{"id": "train_41709", "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: 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": ["ci_integration", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31b866b95d44a92bc35b0911ee5de253b66e0c52"} +{"id": "train_41710", "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": ["reproducibility", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8444e6e6e9d916333b9ceb22fdcbbe8d8cb90e94"} +{"id": "train_41711", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfe2d3148d1b2362ef4db4a067123fd347c9d920"} +{"id": "train_41712", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "546879b328d462051b1b6a35b4023567d2508e5c"} +{"id": "train_41713", "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: 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": ["tests_are_truth", "governance", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "566f3c300bfa13159ba617e595c7eec4b2d8a1fc"} +{"id": "train_41714", "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: 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", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16f33fb57db0d4f2809d7d3a9b214c13856d405a"} +{"id": "train_41715", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6b52e763bea06d0a11516a6a078e958c830a224"} +{"id": "train_41716", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20c5c0c501d9723841dde21892ebc1717d11a5e9"} +{"id": "train_41717", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dc0618b6b80dabb9ef9728b6f5b984854462cca"} +{"id": "train_41718", "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": ["documentation", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef985048ee27a820a8a2856c0e9db98b9ded104"} +{"id": "train_41719", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65203b78129a603510491bacc529f0776179b0e1"} +{"id": "train_41720", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["governance", "reproducibility", "auditability", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "638ebf71e6725a66484408dcc694c7ff1645d4db"} +{"id": "train_41721", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89f7fb12f2e22ce6a33c1af98e73b6939bce68b9"} +{"id": "train_41722", "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": ["security_gates", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57c38ec0d62d59d82d625c84f9355b7de20f46d8"} +{"id": "train_41723", "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: 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", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17e269b3a33a448236f9b9e0c2a79081f8104ae9"} +{"id": "train_41724", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bda778d7be2f5aa8095163bb5cca06f2a83b60e5"} +{"id": "train_41725", "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": ["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", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed4dbf94e228f0243f0f90bf01a46f035c5c5dd5"} +{"id": "train_41726", "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: 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11f0c377ff8d9aab111e60239df3d0163237d274"} +{"id": "train_41727", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8943c336f8efefaa9fe9ce4399428a943d99bd14"} +{"id": "train_41728", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "089c152d02c946e67c49f17a19c4f502f5d0523d"} +{"id": "train_41729", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["tooling", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e69b090b05d7ff9cfb19e4e0183a68c38e5efb1"} +{"id": "train_41730", "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: 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", "tests_are_truth", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9da86a47d465fb779ff3fc0574f7f4f526e7335"} +{"id": "train_41731", "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: 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", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88329af195baff8ce55aa245a2810f68649993c4"} +{"id": "train_41732", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df6fe9b8c9c4d06e2f5bbd3a95090889182013ef"} +{"id": "train_41733", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9df07894a54b0b6484f7286b12437c4527f1537"} +{"id": "train_41734", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6584e2eb6874dbc1497cd5ac777202a76d0adf76"} +{"id": "train_41735", "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: 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": ["tooling", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a83aa3da5d8ff872c3c36495b6ba4a9d89424242"} +{"id": "train_41736", "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: 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": ["reproducibility", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c64c86ce33603fd0905faf66486aecd737002e38"} +{"id": "train_41737", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f21bb6b2c793ce518552fc925987874ab79bf94"} +{"id": "train_41738", "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": ["ci_integration", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34416546579f1149b724e5879670c2363ee988d7"} +{"id": "train_41739", "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: 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", "tests_are_truth", "auditability", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9903ea4ab1b07adb740df25bcd1deeec7e6c28e"} +{"id": "train_41740", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "tooling", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47bbda896c62ac25cb1db4e9e2d77377f382fdd9"} +{"id": "train_41741", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e63d09a7d58e760397b058ded7a490052edeff02"} +{"id": "train_41742", "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": ["documentation", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b68dd0f37c27fcc29356f00ab9206b84ebce4e28"} +{"id": "train_41743", "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: 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": ["evaluation_metrics", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ee3306d68d1b46a20b62d9b9b791b6a25c1f832"} +{"id": "train_41744", "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": ["ci_integration", "cost_latency_tradeoffs", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9e3e8da0cfe7c1b6a326628627506f5e355b223"} +{"id": "train_41745", "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: 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", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad6133ddd9abb2c28dffc47a0b19266c523eb8af"} +{"id": "train_41746", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "cost_latency_tradeoffs", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1a389580ce94179a394494f620aeba702103549"} +{"id": "train_41747", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eab970f17063ea791dfd9529a7c1c1d9f5efd8ba"} +{"id": "train_41748", "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", "auditability", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89fdce024dcc0299dc05a03a0d25247dc93c183c"} +{"id": "train_41749", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ce5463436cb8a3f61ef8d269a16e3b7d0d9993f"} +{"id": "train_41750", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "governance", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7eab55a1d90f8d41b43e093e63c3dccf121d7971"} +{"id": "train_41751", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7eab55a1d90f8d41b43e093e63c3dccf121d7971"} +{"id": "train_41752", "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: 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": ["cost_latency_tradeoffs", "tooling", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7efb8c4d041bf08e6fe9c17d12d68e67df613ad8"} +{"id": "train_41753", "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": ["documentation", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "651527682d2c7a94bdeb1777098c7c73ab939849"} +{"id": "train_41754", "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": ["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", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aed8eb0cfaa56de2938213e3cc554552c8666796"} +{"id": "train_41755", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df4ba711038f071f03297aa99dbb4bce8727af7d"} +{"id": "train_41756", "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: 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", "tooling", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bedae43e68053d57dd9e31bb15eab562e7d41b0d"} +{"id": "train_41757", "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: 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47ee5ec5e6b6c94a0bd017262742ef60ef6439fe"} +{"id": "train_41758", "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": ["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", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "780ba11b521df7e73a97ead9393aeee1444e5e1c"} +{"id": "train_41759", "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: 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", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ae83f84c63a2c152a5149fff8cf22830889b3e7"} +{"id": "train_41760", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["tooling", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9dcf720f523e9a761e0c813d5a9ab087e78158ae"} +{"id": "train_41761", "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": ["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", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bc62326acc5365376bda1670d59403eceaa0730"} +{"id": "train_41762", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "771ead1205228cf62751dd026de7638e8ec328b4"} +{"id": "train_41763", "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: 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": ["auditability", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7eb3470951ee9df997e1825a8ec9bbbbbbbc6cb0"} +{"id": "train_41764", "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: 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": ["reproducibility", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "816c6a7d13668c7902ef6f24fd38924390dd4634"} +{"id": "train_41765", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "112ff6ad8476e8e954ab8ffa67f24bf7c4a4b14b"} +{"id": "train_41766", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["reproducibility", "ci_integration", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e10e5b11cd9d3c5532ebb9b3bdb0ddaf7404a54c"} +{"id": "train_41767", "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": ["auditability", "cost_latency_tradeoffs", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e761b08786917b7a5ba908bbae30c2b23b3ab12c"} +{"id": "train_41768", "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": ["documentation", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4b58af8891e9f860120f66738ecbdeb3b3b30cd"} +{"id": "train_41769", "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: 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", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c75d41f9117a12708f3b0f1f273cf54bd86a7ee"} +{"id": "train_41770", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f8cd9e0292e19e262aefdcbb37a90bdc0a96670"} +{"id": "train_41771", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86d42d17d273ea932fcb7c431041271fd1087387"} +{"id": "train_41772", "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: 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", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d60681b51348ef287729949218cdd60125f6c17"} +{"id": "train_41773", "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: 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": ["security_gates", "documentation", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77fe5a30b14161ce0c9d74ba4bbf1b2cfbb9e1b6"} +{"id": "train_41774", "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": ["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", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcd98baef534d7b653859fbf295a75b9e6113189"} +{"id": "train_41775", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18c19e717f037439136bc2a8d8eaf9437e99e0da"} +{"id": "train_41776", "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: 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": ["cost_latency_tradeoffs", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ee9b87cdb452bacabb7ea948c2da483b8c6f6e0"} +{"id": "train_41777", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready 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": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "848aad42400e18eca3953a1ad5059a881f7ff13c"} +{"id": "train_41778", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2bb098810c2d01ce5082364638b0b107016862a"} +{"id": "train_41779", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdcb4ee342c5833f1d54c9d1e9c2bb12afcfa88e"} +{"id": "train_41780", "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: 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": ["repo_scale_reasoning", "ci_integration", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8994642a323e119c8f57c8eb9f6e3c6e66b5a97a"} +{"id": "train_41781", "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: 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", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cac6f694ad0839d5f8c86c669399d4d8eb80cdf7"} +{"id": "train_41782", "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": ["security_gates", "auditability", "documentation", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ff282b72231711815e0d3eade06df465f1ec9c7"} +{"id": "train_41783", "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: 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": ["ci_integration", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16528b8f5dd93d54657e57d168f5d8b01db8f187"} +{"id": "train_41784", "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: 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", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85476f4c7612f2fd877fcaa4457c66306bce16c9"} +{"id": "train_41785", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0707fa06fa358d37192996ac46eba396e642122"} +{"id": "train_41786", "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: 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": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf8ce836b38a4c74490cf8ba62e0a642ec643273"} +{"id": "train_41787", "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: 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", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7955cd3d63a03a98d55d653814bba7eb06d318e2"} +{"id": "train_41788", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0571532701c083f0e60a8e3d5d3797dcc7a354bd"} +{"id": "train_41789", "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": ["governance", "evaluation_metrics", "tooling", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a876b74dd01011ecbeb163ebc158203e3f92c72a"} +{"id": "train_41790", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1d86aa3edf3535dc070d1ba34f9ac1bb00f8b69"} +{"id": "train_41791", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba2f6293ced69a1199dc77ebba6c69539d430797"} +{"id": "train_41792", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "documentation", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f593188eaf1f3ad4120fe7e8b80387e60fb935dc"} +{"id": "train_41793", "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: 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", "tooling", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6aadce92c9cb280d76b208bcba576f801a681126"} +{"id": "train_41794", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6698350ea0edb082fa5c5faacf0261933d5c601b"} +{"id": "train_41795", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3df658f0160f1aae81aeb5282502ca73aefe5a7b"} +{"id": "train_41796", "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": ["documentation", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1055eba9ad97437efd936a4212791b98883d0a4"} +{"id": "train_41797", "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: 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": ["repo_scale_reasoning", "documentation", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84fd349c9cec62762938b3ec7eba9b6394df2340"} +{"id": "train_41798", "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": ["security_gates", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0b95fdbb49c8d5fa3211b5dcfc5056a7a243a36"} +{"id": "train_41799", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e30027a9304757d4ee801ffb4d5bab3e8fdd4b96"} +{"id": "train_41800", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e08d1a49b5ded90c8ee985e1e57fd2dae08f389e"} +{"id": "train_41801", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c64c29f0fa38ad291684e0fbb4f6e4dcea6a134"} +{"id": "train_41802", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "321b79ed32c900b15ad5049d6deb4d41da1a9946"} +{"id": "train_41803", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64a82012a6fca645fe87c92327e6ca3f334d8647"} +{"id": "train_41804", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3229d8917e4047c74d14afdbfabcd8549afa0786"} +{"id": "train_41805", "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: 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", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c26b05da744f550842b88169082eb993b8cb43e"} +{"id": "train_41806", "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", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f032e0ef0e2fd0dc053d8bf740342137b7799dd"} +{"id": "train_41807", "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: 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", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29e1f95b66e30ea631dd783c87428f670252471f"} +{"id": "train_41808", "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": ["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", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5512608cb18fbdadb3f74d9fba041dd867f5d7a8"} +{"id": "train_41809", "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: 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": ["security_gates", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b56b553f0d0ada798da8dc9171221a613926846"} +{"id": "train_41810", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e8d2ad9cc1bdaf8d984d67cf62b24398d43c3e2"} +{"id": "train_41811", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "734f9c378a0857223050843810823e170ce9d27d"} +{"id": "train_41812", "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": ["auditability", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1535c4fe7009fc7cd8429623b3180d4a3bca7033"} +{"id": "train_41813", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cf7bcf5ac1d9983f76b892b922601940de7a97a"} +{"id": "train_41814", "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: 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", "auditability", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21ffd77ac62a4e320256a35384ce6f807cd637ee"} +{"id": "train_41815", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c82c0835d8984587cd191e92624daf9753c7323"} +{"id": "train_41816", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready 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", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "309644eeb0240e8bfe9193a9dc1247b126704451"} +{"id": "train_41817", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "repo_scale_reasoning", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9827c06b35bd4161421a911040357fac51ca557"} +{"id": "train_41818", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5ab96165991ebc7092e5d6dfd79d019616eaa15"} +{"id": "train_41819", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e546e8f50b831d71b6321a895435702452dcaf4"} +{"id": "train_41820", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5a19c9bf4e53ce977caff1668a807c2d554b106"} +{"id": "train_41821", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["governance", "security_gates", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92994400e6e5d57b097b963e60cbb9a811ecd520"} +{"id": "train_41822", "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: 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": ["security_gates", "cost_latency_tradeoffs", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8f46c7e5f3f6ec9be427b6c6fe1e92893c225c1"} +{"id": "train_41823", "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: 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", "documentation", "reproducibility", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5038067eaa7d3e25f1d54fd08642dc6117f97f4c"} +{"id": "train_41824", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "documentation", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62f34e898772ae3d6b78f21edd3f0d7f0cee888a"} +{"id": "train_41825", "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: 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": ["security_gates", "auditability", "governance", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fae07c998dcd3a04f42d83360229b5bcedf5afd1"} +{"id": "train_41826", "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: 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": ["governance", "ci_integration", "documentation", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73e50e1922562243093d2f2a8a1da93358741e6c"} +{"id": "train_41827", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0463b0c866e29afe58388863659675ff8a3af261"} +{"id": "train_41828", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35eef3c1299f959af9d8774e252a80b568394ee4"} +{"id": "train_41829", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77e5fdcfc290bd6fc6e66f7dc6ceccb345281e3b"} +{"id": "train_41830", "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: 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": ["security_gates", "repo_scale_reasoning", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf2fe0ce2ac94205a9ab99baab499346e41ed580"} +{"id": "train_41831", "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: 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", "security_gates", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93652c91f0a8cca13a10a77285e3e34796b6de62"} +{"id": "train_41832", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5aa95260e762995df665ef0bef039c3bc6a0eb35"} +{"id": "train_41833", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "governance", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "052ef3ec5936d5b5a2e97cd43d7c2ae4cc363905"} +{"id": "train_41834", "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": ["cost_latency_tradeoffs", "tests_are_truth", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f45c15c371664418cd47c3baa6b733255b66a6f"} +{"id": "train_41835", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "reproducibility", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4b0e3c7de89f94c0ce435268f1747c808f46d24"} +{"id": "train_41836", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready 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", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a07737239d0a19495f5da624901a68205366c09"} +{"id": "train_41837", "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": ["cost_latency_tradeoffs", "tooling", "ci_integration", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cde3c74f962e58cf7baaa8d4539ce2c1f38a5c47"} +{"id": "train_41838", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db37162cc15b16e8c56fbe19029a64e284ad0e56"} +{"id": "train_41839", "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": ["tests_are_truth", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8d2bcf123a78e1ca7737c61e3b18d43003c68ce"} +{"id": "train_41840", "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: 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", "repo_scale_reasoning", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08f5781671324437bb6d2349d2cbc59747029721"} +{"id": "train_41841", "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": ["reproducibility", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6a70545ced29ad791d68bb3f3edd229a94489af"} +{"id": "train_41842", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe0d3ff5f9174a35903c2f373ef90a501e7faba6"} +{"id": "train_41843", "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: 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": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d401012124d24e946a14b6dcf999b811c2bbe55"} +{"id": "train_41844", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c84205a12d22a42589883cf0e1955ad6735e12aa"} +{"id": "train_41845", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tests_are_truth", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "531771e97a39e0288dd7e1e0c8875f79297e76ec"} +{"id": "train_41846", "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": ["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", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95fcce900e805b519c48bba2234fbf98ae49e5e3"} +{"id": "train_41847", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d772a6290b841a6b4b4882b6058b5c5fb048a425"} +{"id": "train_41848", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "feab74f6d5acf8a01640e0714ddb649843e3824a"} +{"id": "train_41849", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["governance", "security_gates", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b763e74b2f52a9b5a934d611c64008d5af2b52ea"} +{"id": "train_41850", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2290b9f25e7242f77fde2c48e0a9d42b6b03a0a"} +{"id": "train_41851", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd2e0671324624cd71d61ff66e905274d74c56c5"} +{"id": "train_41852", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["repo_scale_reasoning", "ci_integration", "security_gates", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27b64f9d96c50a81e7ac05fa0bb9b8e5810e3545"} +{"id": "train_41853", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24e44acb884ea7669af23150b5064e3218b08d5b"} +{"id": "train_41854", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b353c2913e5d701928871a8bd4cb8a1a6c4ef16f"} +{"id": "train_41855", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54d5b1460bc596f8fb801ccff71835c478acac0c"} +{"id": "train_41856", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1db2d84d404602d10a6af363422870801b8c03d7"} +{"id": "train_41857", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de29a1c3bc97391a7178cfd71afd255493e8f65c"} +{"id": "train_41858", "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: 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", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2f121b5fe5ff258fbe39ebe1f4caecc55d31752"} +{"id": "train_41859", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tooling", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fafae026f8d06c38b77c64f4b08abe1c35564b07"} +{"id": "train_41860", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca458ba4318cc949c44864500e70bdff3bf77cd0"} +{"id": "train_41861", "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": ["ci_integration", "tooling", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca5e5447cbbdf790849d38141c4ff59f07e5756c"} +{"id": "train_41862", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67a071be227e45399dcb098d9539d7633d5e69d5"} +{"id": "train_41863", "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": ["reproducibility", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "780ceaea8e36251df5cfa77123a5fa9065e5c4f9"} +{"id": "train_41864", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["security_gates", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "862be95e13016eca10b2520f0ec6210a8fb1fcb7"} +{"id": "train_41865", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58d274f1057d70a306b16d7b635078281a0f2bba"} +{"id": "train_41866", "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": ["security_gates", "governance", "reproducibility", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a0614a25ed381745bf60de73d0d2e771aa76f9c"} +{"id": "train_41867", "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: 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", "repo_scale_reasoning", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba8b62eef8a45ea7cb62f02718f44e35d25c631c"} +{"id": "train_41868", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7dc453840417281d6bf5290833b769021f785ac"} +{"id": "train_41869", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "auditability", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1554ad72b1c4012cb316acb1486de92d6badfcc"} +{"id": "train_41870", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd2a7d54895639a8782da3f59c094d1aac9077de"} +{"id": "train_41871", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d197f1538fb1ae8b86a9bedeb088101b12c22507"} +{"id": "train_41872", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "tooling", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30e01f09b241438fcdc05c168a3c03cefdfa8bb8"} +{"id": "train_41873", "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": ["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", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8632cf9c0c95ad8b36eedf01d5e226d2ddaed1ec"} +{"id": "train_41874", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb62b6f8b77c8499107c24d53e56e8818205dff0"} +{"id": "train_41875", "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": ["tests_are_truth", "governance", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c1451c9d1e91b34cec658e06d3c383046f91736"} +{"id": "train_41876", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c0fc400c01e69e9bfc913a1d8304ab8d6dff1ed"} +{"id": "train_41877", "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": ["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", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19dc817ce69bb058b5d473a7e162d04d1eba1c0c"} +{"id": "train_41878", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "evaluation_metrics", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e76dd0632106e71489f6baa71aadd813059fefc"} +{"id": "train_41879", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "governance", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f50accf8b7edc0fd9e507b605ee048bcb3554750"} +{"id": "train_41880", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["security_gates", "evaluation_metrics", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb91e9cc17b75d690bdc40473808ef118719e355"} +{"id": "train_41881", "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: 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", "reproducibility", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ddc46aa09226b978556f61c04a870c24d87c08d"} +{"id": "train_41882", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready 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", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3312ab162ad54ea0323c2a071cd1afa20b0157e1"} +{"id": "train_41883", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "documentation", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0d849a2c09c2c0f185703e128fdc2cc30042c32"} +{"id": "train_41884", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "governance", "documentation", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcee2f409367b7cbe9a7211981e527d639b797fb"} +{"id": "train_41885", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "849f64c1700c22ee2b76add6573e4b04e2cb5d07"} +{"id": "train_41886", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "960882a3ab7d2fe1abfe65b29383fdf4df2f6c4f"} +{"id": "train_41887", "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: 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", "documentation", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "901b23a2e1008658491a07e241352b12c22e47bc"} +{"id": "train_41888", "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: 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", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87b7d38717a9c4000040720cbfae5ce2b5780071"} +{"id": "train_41889", "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: 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c824859951bc8b25d20c3c8d904b0c9f16bd0876"} +{"id": "train_41890", "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: 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", "evaluation_metrics", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44a160c26691805ede6a34172fd1fad10555368e"} +{"id": "train_41891", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3efcfcf28f2aae0287d63e5ea1227ee1ea0049fb"} +{"id": "train_41892", "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": ["evaluation_metrics", "ci_integration", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a22725dfefb2272282ee5325e6648f38dffea9b1"} +{"id": "train_41893", "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": ["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", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb7d66befaa9163cd0e6df596a9375f049953c3b"} +{"id": "train_41894", "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": ["reproducibility", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8daeb0b0dc212ddff55e86db3360c77869af723"} +{"id": "train_41895", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c0473e723c1bd91a2c0038252503ad2641f62a3"} +{"id": "train_41896", "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": ["reproducibility", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0708930087b4e77ae2439ad3cc469b4a227b8e09"} +{"id": "train_41897", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1580c0eb9172b696f450f1f65da8a4511e3b503a"} +{"id": "train_41898", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "916be52fde9073e1b0e772ae49c4760fbf261c9f"} +{"id": "train_41899", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc042cd7ddcb9f4d528973ce8cec0ae9f316e80e"} +{"id": "train_41900", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d31f8b1e52db87c0ef408a68c4182013e63b71f"} +{"id": "train_41901", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f371044665103d7851bfbce6581fe856bde29fc7"} +{"id": "train_41902", "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": ["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", "repo_scale_reasoning", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e617dc467b4db70c50538389f118a23fc8b662aa"} +{"id": "train_41903", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9e355a5a9600c4b053f02fdb0070426237b5652"} +{"id": "train_41904", "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": ["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", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f6359dfd71813231a1369c61f0af31b2b9aa991"} +{"id": "train_41905", "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: 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", "repo_scale_reasoning", "tooling", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b408046b58a5600c253a13d0dba007eeda7da759"} +{"id": "train_41906", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e16848fa009f266879e207b7ec4f6344040449db"} +{"id": "train_41907", "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: 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": ["documentation", "security_gates", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5837f1028c277ce257678d171eaebaca7c45b51"} +{"id": "train_41908", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7af5328ecc381e85f06a9d05b0f70aef2dfd1b4"} +{"id": "train_41909", "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: 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": ["evaluation_metrics", "ci_integration", "tooling", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e2c00c4843ceb175b78b3b6b8f7fd3d21a2a6e9"} +{"id": "train_41910", "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": ["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", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "686f6c335aeb21a8bd58d2380b2249016e8b4934"} +{"id": "train_41911", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eef218b0cf4bd9b59188f5a2768b72ca9cd796db"} +{"id": "train_41912", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa60adab45a5b7f88902cc65d01887d708b5c012"} +{"id": "train_41913", "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: 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": ["evaluation_metrics", "reproducibility", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a448e4fa41cd62a79016c5adabe6d363843c8099"} +{"id": "train_41914", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22d4c3eeafc2510e8a83bc6fb44096f4f828c94f"} +{"id": "train_41915", "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": ["security_gates", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "084bd250e639e3e9ee8714794cd3886dd3426a0c"} +{"id": "train_41916", "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: 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": ["auditability", "governance", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0925e37a2f4706912d5604eac09c756a5b844dba"} +{"id": "train_41917", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "governance", "documentation", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0b1cafddcad3ef915d2dd99265307c529f97158"} +{"id": "train_41918", "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": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d33de789004b348a66d1e8a4070ded90fe841275"} +{"id": "train_41919", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcfd99fdfc37b541c731f7660b4386346a40908f"} +{"id": "train_41920", "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: 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": ["governance", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6099082e5eba6c30801bc00be459ba3e9e07fb8a"} +{"id": "train_41921", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c93384e101bf5d1c4c466568ed44af1284d9d33"} +{"id": "train_41922", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eab970f17063ea791dfd9529a7c1c1d9f5efd8ba"} +{"id": "train_41923", "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: 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", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "633c8fed64b1d09e5703d5bd7c27d7860d037c65"} +{"id": "train_41924", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e257724b8711c50f7db02599a85a82b0c023f1f4"} +{"id": "train_41925", "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: 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", "documentation", "auditability", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38d1a2b080418d6b824199cf8a58cae0b950d97e"} +{"id": "train_41926", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1a6ab57b6965d1f33aa13d44bcb283927103965"} +{"id": "train_41927", "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": ["security_gates", "documentation", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a397bbd66c73b683f25e10c49e6af49741074ceb"} +{"id": "train_41928", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6420b839012d2015b612313e1f0f84ee8860070f"} +{"id": "train_41929", "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: 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cac9e6911ce01ca47632621804540af3c612492c"} +{"id": "train_41930", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fdfd6f7b1f825182787c9cef7c31ba87a83a54a"} +{"id": "train_41931", "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: 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", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a76cd4910c13c4550660d795de0565c448b21388"} +{"id": "train_41932", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6a32fe123c58a3f9ec7ce33533ad5263dbea699"} +{"id": "train_41933", "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: 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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3462c392b3f9cf70cab0b603d8a9d499a3c1e1be"} +{"id": "train_41934", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53fd9aba28fd70df0cafa344a6c790bc61834f15"} +{"id": "train_41935", "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: 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", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47788c102714eda3970bf1aefd7f1362b450e2b7"} +{"id": "train_41936", "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: 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": ["reproducibility", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ca069cc50fdf8ef9be993fbab584b92e6143f2"} +{"id": "train_41937", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15661ae8085f8278df98a17547507cc30a2db2b5"} +{"id": "train_41938", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c6fe787b255e0e329548c0b5fc0d97339f63e63"} +{"id": "train_41939", "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: 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", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18ca13f5d25697af6402c8147ac6f9f3c39d637b"} +{"id": "train_41940", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90a31dbe3f1451114046b5939921aaa595b2ae34"} +{"id": "train_41941", "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: 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", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "439e16399ee77782ebd609bf4c947bbb7387f949"} +{"id": "train_41942", "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: 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", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92c6319612594ea6bc781b563e5bee470f104878"} +{"id": "train_41943", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d2f24dbc5f59788b75ca5b2cb0895a6eb885c01"} +{"id": "train_41944", "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: 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": ["reproducibility", "security_gates", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44184bbd035c438c1991338d7a82719b05775e16"} +{"id": "train_41945", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "831b82779e9a316e033b1c1697feeda035a70689"} +{"id": "train_41946", "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: 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", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af696000de0bdf18c13c2f3cf41fe2157ecb1cda"} +{"id": "train_41947", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "ci_integration", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cad0f127ec88f14602d71f89ea3100836826185b"} +{"id": "train_41948", "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: 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": ["security_gates", "governance", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3759674716b4bef1c5d64e7e427b521852bcd38"} +{"id": "train_41949", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "security_gates", "governance", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0054961b55b4d7441c02f9556ee881bfae5a6a0b"} +{"id": "train_41950", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d11301a105adc727e13935aa8ff11a856c0b63f5"} +{"id": "train_41951", "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: 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": ["tests_are_truth", "ci_integration", "documentation", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2b3c1fb17628ea819a697d66f9294a7d06680e1"} +{"id": "train_41952", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "021c7480b31533d9032c716820212d77e5b8a09b"} +{"id": "train_41953", "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": ["security_gates", "governance", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2316773b59921a6b4f068bcafea4895e2b49b59b"} +{"id": "train_41954", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e648df8cdf5eb288e661db08a730d0084c84b7c4"} +{"id": "train_41955", "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: 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": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b712e5c0b21f8a9f2909a816ba318cb29814770d"} +{"id": "train_41956", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "documentation", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70a7cd6b8a42353aadf9d89ca753c88aa61e5e51"} +{"id": "train_41957", "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: 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", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e741f1662246ce3e762af07dc2fafce91ac8e2b4"} +{"id": "train_41958", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "719e3f83be77eb463be2442c37d838f3ca7e8c91"} +{"id": "train_41959", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["ci_integration", "evaluation_metrics", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fa463b899706d660c4c73d875476c3f09f29176"} +{"id": "train_41960", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "370931a7ebb3937f80647ba2df1087af7b0b25b4"} +{"id": "train_41961", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49fa57d269a3fb86b1cc6c7d11831c3b2a9fa9c7"} +{"id": "train_41962", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb26d8741fe77855ad6427dd0ab8a004fb6bc195"} +{"id": "train_41963", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8052f26f2c0cc4af741aabc6dc902644cfb04e45"} +{"id": "train_41964", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "documentation", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "133ee636bb85b1697cca80272c3a367a66fb0d3e"} +{"id": "train_41965", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e86b4e2a9e5f0041465e9bfb5182bf04042630a9"} +{"id": "train_41966", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tests_are_truth", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d070d62c9e108be72f5671954b532242dfa7ec2f"} +{"id": "train_41967", "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": ["security_gates", "governance", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07b71f88ea4ca89d74fbbd7389703b323f7847be"} +{"id": "train_41968", "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: 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": ["cost_latency_tradeoffs", "governance", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "703205ebf04858526f7a6f2c59e92cf709359e9e"} +{"id": "train_41969", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3106c5fc4385724208bf7555a10318c42aeb1d8c"} +{"id": "train_41970", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e57a995dc865bae523bf35e51d82a13dbf56b3f8"} +{"id": "train_41971", "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: 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", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4a22f19a6be7d35e9646b8a8e4f956c9af0ae72"} +{"id": "train_41972", "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: 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", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f67f8c0969aaf95fe750dc4401855deba215fd64"} +{"id": "train_41973", "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": ["evaluation_metrics", "ci_integration", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2017d1f1ab3104f75c7c6f1b4a2d65dd8a11a0c"} +{"id": "train_41974", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1da9e360894aa68bf3ed84177a7304cb84af4e5e"} +{"id": "train_41975", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "179287bc14e9233c057ad5706f440ad22da7c744"} +{"id": "train_41976", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f81cef50d0962ec1828ceed32be6585aafbf9433"} +{"id": "train_41977", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5c7f12c5cdc28fc7e190a4a8884b1b9abfd46b8"} +{"id": "train_41978", "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": ["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", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ca8ee6a4787745cf1693955ac040ca507148770"} +{"id": "train_41979", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5afbd6379237ed65e9bfa1afe0c44c269e2b9de"} +{"id": "train_41980", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18cbf4ddfba22679c62e76bc7b2397813c621f1b"} +{"id": "train_41981", "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: 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", "tooling", "security_gates", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dfa7fb1f7c184edff2ed0a492734ca0d1e5bf34"} +{"id": "train_41982", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52dfac0615b9ea51c8943c153a5690e144a56f8c"} +{"id": "train_41983", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cafadcda85649d9fdadaaedb7fd0b2a1dff2194"} +{"id": "train_41984", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62622fca2911a58f244aeaabe18f06439a2be1d3"} +{"id": "train_41985", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20b0a6f94102390dbd8a256c2a6568816c20f4b5"} +{"id": "train_41986", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e974df8e1d7ed2513c8c6ddc631b7ff2baae2500"} +{"id": "train_41987", "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: 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", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8e585307df0095c6e25ec5c47b1b7320a4d8161"} +{"id": "train_41988", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["documentation", "governance", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7363c627c1f7687b1b5d0b3fb43ea2195803be63"} +{"id": "train_41989", "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: 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": ["ci_integration", "documentation", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2b3c1fb17628ea819a697d66f9294a7d06680e1"} +{"id": "train_41990", "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": ["repo_scale_reasoning", "security_gates", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cac5ebd5d0795e1e559bba0fac54c744da1d1af4"} +{"id": "train_41991", "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": ["tooling", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "646f834054ce7d2605f473272297922a8b4ebd74"} +{"id": "train_41992", "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: 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": ["auditability", "tooling", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4a53a33c5a5aee2ce8aadc10c1607056579da3e"} +{"id": "train_41993", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccc727e7523625114a2b47edb7cfaa4e8006b64b"} +{"id": "train_41994", "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": ["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", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30d5b412b52e4933b1f2158e5b75479d7e4b729b"} +{"id": "train_41995", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "documentation", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4904c4351679a8dc21a0ebd86bf0c59441b216a1"} +{"id": "train_41996", "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: 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", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7858744b5023807e551a9dd4dd539f421fdf80a8"} +{"id": "train_41997", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64937b8bc9491f5f1fcf3bab915c3332b955dc87"} +{"id": "train_41998", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "207322ac17b4351ab1a1803589309e47138e6c18"} +{"id": "train_41999", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94fa0cdbc025fd331ebe1afa23ee81c191789fc8"} +{"id": "train_42000", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87a5263c3f1344c6a407aabe57d92c15c43a8f03"} +{"id": "train_42001", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "484f6eaa79e765e4a1f5aa9223744d1f80aaf3e0"} +{"id": "train_42002", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f214dad4a53352219b203e056a18c2d8f7942e11"} +{"id": "train_42003", "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: 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", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04401ca1cd0baa2f80afffb89fedb824d9fbd529"} +{"id": "train_42004", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "governance", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74950df59114b75a8273ddf9b1686cf9e1316486"} +{"id": "train_42005", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04e1850e2132a168069571c699f41bd9652058b4"} +{"id": "train_42006", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c98b86a994e1cd7cd8d456f2a936b080732b1c4e"} +{"id": "train_42007", "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": ["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", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05c64e7683bdf194de1755f987c8ce54d62f345c"} +{"id": "train_42008", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eae855e82252b0268fedf0b80b0a4f3e70dae5f2"} +{"id": "train_42009", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "befd1a16a0696ec1e2c2957d27c568278b754a0d"} +{"id": "train_42010", "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: 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": ["tests_are_truth", "auditability", "reproducibility", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58e60cc0defff9832fe277b7242ff0ccb6728aed"} +{"id": "train_42011", "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: 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", "ci_integration", "tooling", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecd25a74d3e318ff3cdd2312a64d8d123c93fed8"} +{"id": "train_42012", "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: 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", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2675ef3bda42d345643a9a797597405cc41ad440"} +{"id": "train_42013", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9faf4e3d7876d89b5a829cf5ba31e14e6d88a46b"} +{"id": "train_42014", "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: 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", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b03a42f233c2815a065acdb74649cb1b26000954"} +{"id": "train_42015", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f9583abec7a9b05f63041b628fd4341ff5d940f"} +{"id": "train_42016", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "tooling", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09d65e331e02f647b74923a927893382fe191844"} +{"id": "train_42017", "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": ["cost_latency_tradeoffs", "repo_scale_reasoning", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6202d627269b367468269a7b716e38f62e27fdf3"} +{"id": "train_42018", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec2b7632c5af7653ce0f946a2994e920a78fc038"} +{"id": "train_42019", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4a13ab09a9df5f45bf5289e58abdede4e5109ce"} +{"id": "train_42020", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["documentation", "security_gates", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d1fb465d0fd1e692a7b53979f3df16b088f131f"} +{"id": "train_42021", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1062ca2959dd0e793a24fd4eb8a43f392496262f"} +{"id": "train_42022", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["documentation", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23a6549553e3d1fb8c483fc3ac801cb33797f4fa"} +{"id": "train_42023", "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", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb2ff79bf0d5c2b1a0d2db702d8fde183fdae6ed"} +{"id": "train_42024", "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: 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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62552cf96d8876928cdd022192f217836da40ea9"} +{"id": "train_42025", "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": ["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", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ede9b8d2f505f57a28bfb719b9a5baaa70b1c96"} +{"id": "train_42026", "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": ["tests_are_truth", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25c3c312d04b5b3c5f9777dcc8da6522020e26cd"} +{"id": "train_42027", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47205dcd54583344d1401cea39ad019c6367a5a1"} +{"id": "train_42028", "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": ["tests_are_truth", "documentation", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16ed11c2e575d4920e9ad4011dc04b2afe44817d"} +{"id": "train_42029", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["tests_are_truth", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a53150fc9fea421c3bef73e76f9adf88d5944ca1"} +{"id": "train_42030", "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: 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", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c996019e77cb745c67be7bac2c4ee00f805cfef3"} +{"id": "train_42031", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52c3764075dc36eadd3397ce75c130bad4953019"} +{"id": "train_42032", "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": ["evaluation_metrics", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9635e598182370cc87840065a4a289773177e48"} +{"id": "train_42033", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03a7170135032cff36624eeb21981cd1cafc1a7b"} +{"id": "train_42034", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "831307fe7338affff352a194811d29ac9183a032"} +{"id": "train_42035", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a5121faa79310d7b5a58cc758560e493f547227"} +{"id": "train_42036", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "auditability", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12cc973fae66cbe81dad7a58394e5130673b924f"} +{"id": "train_42037", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "security_gates", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3661cdbc54b7cb017aa0a2812baa0e2f83f1f0f"} +{"id": "train_42038", "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: 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": ["auditability", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d93c9a556a8809c750426d97d9c05c49730a209"} +{"id": "train_42039", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ee4152892bdf7db648d2e40a249bd884fb10fbc"} +{"id": "train_42040", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06356be88032a7c0cbcd1033438c4efec54b82bf"} +{"id": "train_42041", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["reproducibility", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d8229aaa23d817f5a7feccb1c4c83f36b2c64ac"} +{"id": "train_42042", "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: 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": ["evaluation_metrics", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf0991058eef10b482cae76b7ebca2bddeb7d69c"} +{"id": "train_42043", "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": ["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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8fb131c42fff326761891678cc632aa24a7aa5d7"} +{"id": "train_42044", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecafb6cd0f25506751884792e634b5fe977aabd6"} +{"id": "train_42045", "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: 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", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a2588269447f20bb487a800c808e5f184bdf4a5"} +{"id": "train_42046", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6dd67450c48ee6179a3d02ec3b2d1a6964b2862"} +{"id": "train_42047", "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: 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": ["security_gates", "governance", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a20ecf5c3698f3ff6e763cc71ab391f5a03990e5"} +{"id": "train_42048", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ce23e54c276bc7aec49fe40f4ae61bc298c9d7f"} +{"id": "train_42049", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "059b9cb867b8caaea0594431ab041c70a091171f"} +{"id": "train_42050", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "documentation", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a7d117f046270288f416902d4cb333f0f7fe9ea"} +{"id": "train_42051", "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: 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", "reproducibility", "governance", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "926ecebd817202758ab453bc860de3b2d31bf007"} +{"id": "train_42052", "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": ["cost_latency_tradeoffs", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5935622e0f812762e75a52b4d273f52e862938a"} +{"id": "train_42053", "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: 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": ["repo_scale_reasoning", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5b4ede8b08424002a7aad29eaf31295d9cbb16b"} +{"id": "train_42054", "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": ["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", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "482e1c978a898ecad89b70adfe5a74b1028979fd"} +{"id": "train_42055", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10bd1554a514ea7b876b18bb892858f57f313e30"} +{"id": "train_42056", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65e9e75812d0d84f2130f3668e1a6e2607330951"} +{"id": "train_42057", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1c6b3fd8ea4b966673e13a5f9dbf45dd8f8f36b"} +{"id": "train_42058", "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: 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": ["tooling", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a19f27e125522fad1fbbc9a648aa169c11da0a44"} +{"id": "train_42059", "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": ["security_gates", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84aa9b68e3b9f00f17645fbeb3f0a46594264ada"} +{"id": "train_42060", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "governance", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13d5be8ec803db1d5580f03c87de45dd50666949"} +{"id": "train_42061", "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": ["ci_integration", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "399dfa3a2a71292c6ce9ba7be3954762ccd46563"} +{"id": "train_42062", "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": ["reproducibility", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f50accf8b7edc0fd9e507b605ee048bcb3554750"} +{"id": "train_42063", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02c1ffebfff5c97b1efb71334dad4cb2d5b4d865"} +{"id": "train_42064", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca7011b050cb9f87521f050b3cbf6134c79eb1a3"} +{"id": "train_42065", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70afb55cd8c5d6f984f1e84907c151918c391adb"} +{"id": "train_42066", "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": ["tooling", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "976d21d3ee75ce7f17c61255223052609f0c4326"} +{"id": "train_42067", "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": ["auditability", "tooling", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a265d862bb6f00b5e8e0d5208c6087785b0e12f"} +{"id": "train_42068", "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: 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": ["evaluation_metrics", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7205a92af873f680a139db18721dbccef1c1bec4"} +{"id": "train_42069", "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": ["cost_latency_tradeoffs", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f4c2ce6f327a8aae1e111a83fca47c5f1cad05f"} +{"id": "train_42070", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f3a73c0e88f5e1b027684e3c5efabc39deecc6c"} +{"id": "train_42071", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7561eb37ab066de9c543d9f2df9d8b6690d2128"} +{"id": "train_42072", "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": ["documentation", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31d961f4d27ccfed4a2b193f59f2747d1dbe8216"} +{"id": "train_42073", "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": ["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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2faef4557781a5c3236709ee6f8a94352f04a0e3"} +{"id": "train_42074", "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": ["reproducibility", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6268cc4f1d8cae312b2534aeb426b5dd24bc434e"} +{"id": "train_42075", "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: 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": ["tooling", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b294b0280e020eea1b90b743814cb1311508a308"} +{"id": "train_42076", "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: 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": ["security_gates", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dff4180bbf372f40ca4376fff7143fd9e585d38f"} +{"id": "train_42077", "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: 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", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e637184ac32bcf5f6f26f6471851f5e6e68191cb"} +{"id": "train_42078", "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: 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": ["governance", "cost_latency_tradeoffs", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2304dea39fc90a2d1f1019e48d3501fae3657a16"} +{"id": "train_42079", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40309b458e48fd8fccdf045fca3daffae7e2adca"} +{"id": "train_42080", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc5b2418575706898dd84aa1a8a55cbf71e47e60"} +{"id": "train_42081", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26e17f2142b2851934bfe0522a5860b12c6843ed"} +{"id": "train_42082", "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: 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", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ff3cc96192a9cce66d5ace1103405eb2d8f9c9c"} +{"id": "train_42083", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67ab2bddf4cc7f0103de16c5242c089505113c5e"} +{"id": "train_42084", "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: 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", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cd5b0999f27497962e5e4d23d9682a1e2f84583"} +{"id": "train_42085", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4571f25fefa3e4567e33f99471f07e952732b211"} +{"id": "train_42086", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c29788ad2991b97036f02cde2902e86ea17c096a"} +{"id": "train_42087", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7158c0bd4df402b8c8a1b99c5b215f6db20690ff"} +{"id": "train_42088", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49cfe255142361260891ad0fd9e20dbdec822768"} +{"id": "train_42089", "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: 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", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f2bfb67f1ab8af5fb0de81d22decae5c5cea141"} +{"id": "train_42090", "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: 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", "reproducibility", "security_gates", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff64a2e1e6a1d66315344ba6daceab31267986bc"} +{"id": "train_42091", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6ccc2c4b5b45d517cf121f7ef9be559d67f5be0"} +{"id": "train_42092", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a8ac2e5a0070e4bf48745edfc9a7ab1b57257de"} +{"id": "train_42093", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "governance", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "291c17dc59fd3e9478556fc873527f0e9e09bd76"} +{"id": "train_42094", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec59e3840ced6b083b9fd24103d962f3dc6ed617"} +{"id": "train_42095", "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": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f190cd0d9c1359f868e81f1a928c807c1783fbe"} +{"id": "train_42096", "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: 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", "reproducibility", "security_gates", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfd4e56c9ef653951a9d7c28eb003e931e38b813"} +{"id": "train_42097", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6baa85b6ad048998c451b208da6f9246f2add6cc"} +{"id": "train_42098", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be32227ccdc89d651054825ff9926898f37486d5"} +{"id": "train_42099", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "504e39daf72f4d22cb281719158652fba766ad7a"} +{"id": "train_42100", "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: 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", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4651db852b609dd38ea5bfc91f90d0b480ffeffd"} +{"id": "train_42101", "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": ["security_gates", "documentation", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4866435fe9d945a093c55fabd1f144bad7715c66"} +{"id": "train_42102", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "101b10f131045dfdf48c85a2d3fca686ba54bd37"} +{"id": "train_42103", "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: 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", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "215b0a8e0393adbbda41529b4f09ce711862686a"} +{"id": "train_42104", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08a9f5b4f5cff381f8fecb61a361850014227c4a"} +{"id": "train_42105", "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: 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", "security_gates", "reproducibility", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee3b7ab01bda7b5a3214b0c03e43fff90db9f9e1"} +{"id": "train_42106", "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: 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": ["documentation", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3c4d68ba32dfbabd4d245d8ca5e6c2489730d15"} +{"id": "train_42107", "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: 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d63bac7f04867d5de5bd1cd5b35b712fbadb79e3"} +{"id": "train_42108", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0b54ed1c28d41e0748c4ac97c9b9abc3cfc2a23"} +{"id": "train_42109", "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: 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": ["auditability", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c99a894443be9c3024ccb45cd7f932a857616d2"} +{"id": "train_42110", "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": ["documentation", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "725f94befe7296127ac2cf81472451faa6f07212"} +{"id": "train_42111", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3ae75400d50d4860bdf69769bbb28550c4b774d"} +{"id": "train_42112", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a06b9eea8584902a63ff9151d9a695570b60ed35"} +{"id": "train_42113", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b8fefdf833d0e6cd9887cc358f724bc5675fe7c"} +{"id": "train_42114", "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: 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", "auditability", "security_gates", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46bf75f5c75be4178f4dbfa9152aa738e25b1263"} +{"id": "train_42115", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0c7f08afef7712b913c7fdb44e35e295940fb92"} +{"id": "train_42116", "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", "tooling", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e18904e3eb400b0732a00256ea923b3e88a030c7"} +{"id": "train_42117", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e8d2ad9cc1bdaf8d984d67cf62b24398d43c3e2"} +{"id": "train_42118", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["security_gates", "documentation", "tooling", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95b1435a52b0cc9e67de0a4f6ad5f03e2bd76539"} +{"id": "train_42119", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "084fcbfbc11c8e08935b9788e1a80fd792ebaf7c"} +{"id": "train_42120", "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: 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", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47b39ab29f9afcec2d53f12282ea853f116f5912"} +{"id": "train_42121", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cff003dbab5f1aeec44c12c761c270d4fe9e694c"} +{"id": "train_42122", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "017814982591252d222d58844346372c5f6199d2"} +{"id": "train_42123", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e60b7c25c33bb805c2d5cdf0a135f43e3fea1ff"} +{"id": "train_42124", "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: 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": ["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", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "576ec2fe3f6a0487e9a9110b1ce953b49969bb68"} +{"id": "train_42125", "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": ["tests_are_truth", "tooling", "security_gates", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41da74e3dac5f6cd1d7d59afb72b8ed8a95d7016"} +{"id": "train_42126", "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: 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": ["reproducibility", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "986fa607ac2986c771a312615e1ee3a734745917"} +{"id": "train_42127", "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: 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": ["documentation", "repo_scale_reasoning", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34cfe6bdbfb41e5f99a6b5405f311c2976d84221"} +{"id": "train_42128", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d1b6335176a96afd72061e12dc85bbe988879aa"} +{"id": "train_42129", "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: 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": ["auditability", "reproducibility", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3dbef86dce2d2ffc0fc0048e13248976828a07a"} +{"id": "train_42130", "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: 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", "documentation", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "919f8e8a825204f9e680c9382eb80052719e4475"} +{"id": "train_42131", "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": ["repo_scale_reasoning", "auditability", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ac128bd1a493b3b76afbc509dca25a4ab36b449"} +{"id": "train_42132", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad29fa2549c09fcb18f32574166ee49c474e4673"} +{"id": "train_42133", "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: 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": ["governance", "auditability", "tooling", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a1a907219ff106592f1ff414cd6fe9002012729"} +{"id": "train_42134", "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: 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", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4200f0ffb83ee26e1aeb1711332a7c46f4b318b9"} +{"id": "train_42135", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "943dbc514f5a6e2b7916e8aec4e27abe356fcee4"} +{"id": "train_42136", "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": ["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", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "644f43bfa5f169b775a6d7e565a63bbaeaec811e"} +{"id": "train_42137", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5f3e85b9adb9698404bf6f19a027a153f522df8"} +{"id": "train_42138", "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: 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": ["documentation", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7533aee336226fa14141482f41a1c2d27b6e085"} +{"id": "train_42139", "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: 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", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1a57b97529ff0aa96e152049b448d02198cdba6"} +{"id": "train_42140", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab7ffacff5e2f307fa820066718a7a364f606d3c"} +{"id": "train_42141", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3947b3a532c2f98acd4ae00cc72ee55fffe83dbf"} +{"id": "train_42142", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96628122349554184ddbf34934e53ecb61f791c0"} +{"id": "train_42143", "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: 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": ["cost_latency_tradeoffs", "tests_are_truth", "tooling", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e89ceec56154563183f75288ad86bacba097687"} +{"id": "train_42144", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60a9610921d20f783e48d8433cad2e9441625062"} +{"id": "train_42145", "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: 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": ["repo_scale_reasoning", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a713524317b97115197d8f1e09946ed95969bfc"} +{"id": "train_42146", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b8385034f99bcf103606b5419a46615da443cc9"} +{"id": "train_42147", "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: 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", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92eefd5dfc2adb431d7060587785a2ba0c87382b"} +{"id": "train_42148", "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: 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": ["tests_are_truth", "documentation", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb6edc24a3f2e57f8e0b8f076e871cf69d3cfeeb"} +{"id": "train_42149", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "governance", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d5e95eafe4cdfcdc369a052aeefd2483dcfb8c8"} +{"id": "train_42150", "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: 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", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f785220e091cf1c244d216868f172d257938f50f"} +{"id": "train_42151", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63c2e33f188cf622836a4814ec2f7f055cfa8f42"} +{"id": "train_42152", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78a4aa99c55faf3898bdcdcd8d467e6ed7aa0fd8"} +{"id": "train_42153", "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": ["tests_are_truth", "evaluation_metrics", "ci_integration", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdde8b5c36cb59ae1f692edadb3dd9a16b977780"} +{"id": "train_42154", "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": ["auditability", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b644826a8eb827f86350783e8826beb38763e06f"} +{"id": "train_42155", "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": ["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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ca410e5c81e7ae7912d15d01d838e0ee541e0a0"} +{"id": "train_42156", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b63ffafbb4aa0426dacb4f449f8cfb7a9f22b2e"} +{"id": "train_42157", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["tests_are_truth", "reproducibility", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab99c0e28d59690bc112c4cf30d2380da4013eb4"} +{"id": "train_42158", "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: 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": ["evaluation_metrics", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd964af229b0cd13dd91d56877272b5f36f71362"} +{"id": "train_42159", "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": ["governance", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23d9a00981708ac705d6c4d68bdc87bfdff4a6f9"} +{"id": "train_42160", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ceee9cf8dc2230f08a93b2142324d5065accc2b"} +{"id": "train_42161", "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: 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", "reproducibility", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f29fd1df44bbf61453c9a64e94b9c18f45cf1b0b"} +{"id": "train_42162", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09576fc11b40d84526b6a8b14e9bb63e04f91adc"} +{"id": "train_42163", "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: 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": ["evaluation_metrics", "tests_are_truth", "governance", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78fff19d86978f748bc2272a8faebffb48797e1a"} +{"id": "train_42164", "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: 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", "reproducibility", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf6003dff70abf96796778c9a6724cf3a6e4baa7"} +{"id": "train_42165", "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": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc0c2ee21295ca345dd0ee558e75d8fd41f9ca78"} +{"id": "train_42166", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35abb5a4a5a59bf98c14cafcd6fc02afd21eda05"} +{"id": "train_42167", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c6ed7af229fbea693d7dcb0eb3f98cf2b03a909"} +{"id": "train_42168", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "auditability", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "438e41be4677cd716440875351af88a91b5c36be"} +{"id": "train_42169", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8b275459e9d048374e19bf8b53126c889036d41"} +{"id": "train_42170", "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: 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": ["evaluation_metrics", "tooling", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5afe4a0edef47555a07359f9cb562052a6c12736"} +{"id": "train_42171", "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: 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": ["documentation", "reproducibility", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e98aac2429ef79086e2089089d2be044a00af625"} +{"id": "train_42172", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45db66df989765e45271a699ec272d6b69ba17f9"} +{"id": "train_42173", "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: 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", "documentation", "tooling", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "904aca0b3a9f5da2ab9a42e26ab67067d4c5f554"} +{"id": "train_42174", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7be1ac5ce4abe9971c77becab0111cdbc5025613"} +{"id": "train_42175", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e9e99679dc32584ae0605fe023e63d52bffcd01"} +{"id": "train_42176", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fdfbb0ba4fdea10a7471bb2b1c93979bea640f3"} +{"id": "train_42177", "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": ["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", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7ff266eb60b027d67395bbab8b0e7632e6757ac"} +{"id": "train_42178", "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: 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", "evaluation_metrics", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cc759e80f5428f5a4f125614563b6040970ea32"} +{"id": "train_42179", "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": ["security_gates", "ci_integration", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8702533e65c9e003e67116834e63bbde136fe94d"} +{"id": "train_42180", "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: 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", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6247c587df795c5c2c4fbdd0029a13c45c67f105"} +{"id": "train_42181", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37de0f5518f02b8c861bf77e9eaf0f0889848a6f"} +{"id": "train_42182", "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: 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": ["repo_scale_reasoning", "tooling", "ci_integration", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "768e5f93953c7b73f127110e767f65998034c4eb"} +{"id": "train_42183", "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", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b7404dce9dd5b1e5a21da7e61a8612aa4c074df"} +{"id": "train_42184", "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: 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", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "798aedff363a2bf6a6821c1b1d7c0714d6d5d3ba"} +{"id": "train_42185", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "documentation", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f27fdf6511d049cbc62aa177a76974b5df655686"} +{"id": "train_42186", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["cost_latency_tradeoffs", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2fb2af9b00b0e74b6a37db50c69db42a34d1e0d"} +{"id": "train_42187", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8052f26f2c0cc4af741aabc6dc902644cfb04e45"} +{"id": "train_42188", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1da9e360894aa68bf3ed84177a7304cb84af4e5e"} +{"id": "train_42189", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3aa6afaa163b3df5914f8afc9324d2bb216ad644"} +{"id": "train_42190", "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: 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", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c6f651767365db0bda4887959c5bea19d047f13"} +{"id": "train_42191", "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: 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", "governance", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5d898966be5477e6be0b9b37edccadfc4dbdd94"} +{"id": "train_42192", "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": ["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", "governance", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "977712837ed496e49ce7b75ca41f787b73407c67"} +{"id": "train_42193", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "257b7933e8d3e0974987034be8c0941c515a3544"} +{"id": "train_42194", "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": ["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", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62936cc5ce4ad14fd71627f11c8a4ee736f6d830"} +{"id": "train_42195", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16e4970ac029f15f85f8cd33e7c786eb4539e592"} +{"id": "train_42196", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec2b7632c5af7653ce0f946a2994e920a78fc038"} +{"id": "train_42197", "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: 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", "governance", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "618c7fb26b8bf5b487fa836241050c89587531ff"} +{"id": "train_42198", "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": ["ci_integration", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7a3031d97220ffdfdd40e4c7032768a14ac83d1"} +{"id": "train_42199", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e06e710d54b6ebb9658621fc7bd4a8209fbd6752"} +{"id": "train_42200", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9d391edaf886bdd3167de0a15e5f6f8704f94be"} +{"id": "train_42201", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["security_gates", "tooling", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "156e098190817b90aa0bd1a1e7c0ce38dc690c3a"} +{"id": "train_42202", "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: 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": ["repo_scale_reasoning", "reproducibility", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b7a249fc8371a8d6bcccd52aab2980b2b6d22ae"} +{"id": "train_42203", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82ba36cc9b301f49a28186be565a8b52d50fb76f"} +{"id": "train_42204", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c935203403d3adf3a2e04c5d5770e18746aefaf"} +{"id": "train_42205", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e292a0e87216cdc608e6b0e088e3411b7b2745d"} +{"id": "train_42206", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "223390c2e7c6908a92f32a9e0cae55532ced69b3"} +{"id": "train_42207", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "auditability", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8764705f751db6c529c210f10fca2da487e4c93"} +{"id": "train_42208", "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: 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": ["governance", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d93c9a556a8809c750426d97d9c05c49730a209"} +{"id": "train_42209", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d63c1995db608a27248740e5645c1110516af300"} +{"id": "train_42210", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41935207a7ee03eaf7a5819122219224b3f532d3"} +{"id": "train_42211", "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: 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": ["reproducibility", "evaluation_metrics", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c34933fa9a53a08f4d4772498580e4fe8667274"} +{"id": "train_42212", "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: 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d05cdea8f35f77163380f9d1297a6612d8009827"} +{"id": "train_42213", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "229ce157193cfd82efe56f041f4f43959db010d0"} +{"id": "train_42214", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "security_gates", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65a90dbc3d71b89d8bb82c34e2508cc2b48e4552"} +{"id": "train_42215", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83bd8b018e3f13ead1a4d95e1ee6b605c233096f"} +{"id": "train_42216", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9728cddba511b9d20cfc9571cee360cc519373ea"} +{"id": "train_42217", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tooling", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e061d24196a68116ef228e3b517ed9b476cf1037"} +{"id": "train_42218", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b161cf7d9915cb773b5f0e40b5f2e38a0607c12d"} +{"id": "train_42219", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19aa128aaa4ce087e08aee9975017b17618582ea"} +{"id": "train_42220", "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: 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": ["auditability", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d51c7c9607d4161608f0b30bef6033cc550bda59"} +{"id": "train_42221", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12f14dbb7c5d5646e20e6c2e98095fe19c287ef3"} +{"id": "train_42222", "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": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74d9ee14df871cf632841b6df5ea0c0dda0a8a48"} +{"id": "train_42223", "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": ["ci_integration", "auditability", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e3fae20c30ebd2c974c6b00365c6ce4fbd882c4"} +{"id": "train_42224", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c49feb6c42db3aa08b64000ddc3b101097fffeb"} +{"id": "train_42225", "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: 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": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db5b8e8f14ebff7427e33763bdaecbffdd487cf1"} +{"id": "train_42226", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b48c11346f4769a1f46c544621cc4fd5c08b3bf1"} +{"id": "train_42227", "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": ["security_gates", "auditability", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e846b2423384cbb0e998dde282c0135bcef00c5a"} +{"id": "train_42228", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2114ada675b3361010637de370865b62d2fa9f85"} +{"id": "train_42229", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e118f0f919855c1916d4cdf38affa703fc897ff"} +{"id": "train_42230", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "reproducibility", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ad5932b3fcc689de893f63b2027ffb6f67b2cbe"} +{"id": "train_42231", "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": ["reproducibility", "tests_are_truth", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b74647d28e07030c298417115b9bfe7ca6c23f38"} +{"id": "train_42232", "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": ["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", "tooling", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05c858eef2dbddd0765ec8d596611489da6db91b"} +{"id": "train_42233", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08d21b3e5456356cf04fd4e3527700eea5d55af3"} +{"id": "train_42234", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9166a02f06ff96b55a66885a4cfcbd0cba445ae7"} +{"id": "train_42235", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a53d5a64f9cc76e9f59fa72f69a595126311d3f"} +{"id": "train_42236", "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": ["cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fda755bd8a14cdbc4730c7e1c533bec6fd90a476"} +{"id": "train_42237", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dc0c984af7fd48b802945741d56baf05b7d942c"} +{"id": "train_42238", "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": ["evaluation_metrics", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42bf0b6efa0a1261e47fc2d59f52245069189811"} +{"id": "train_42239", "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: 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", "reproducibility", "evaluation_metrics", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7a4a77f75bd29a7b07995fa6c904679d518b250"} +{"id": "train_42240", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "evaluation_metrics", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89f68247ec9d6245baf93506b4e029843c97f53d"} +{"id": "train_42241", "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: 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": ["reproducibility", "cost_latency_tradeoffs", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8667627c99ac36248210b22f79ade54631f9bbb7"} +{"id": "train_42242", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "690b0a4b256afeaa8a702384a50eb075e7d0292a"} +{"id": "train_42243", "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": ["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", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85472f62bb83733a0b98b8bed7cee281bb43dbeb"} +{"id": "train_42244", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["tooling", "auditability", "tests_are_truth", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c98ac729f5403de13ee616a8a20bc06a4edb1657"} +{"id": "train_42245", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "repo_scale_reasoning", "tooling", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b30cd29a83d697669e6019dfd4bf79ab5871731"} +{"id": "train_42246", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92fee0ba901177d655039a5c70c6fc9c024a5937"} +{"id": "train_42247", "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: 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": ["evaluation_metrics", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "493d0d5dafb54380d9d408b2f1b731dff727e4fd"} +{"id": "train_42248", "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": ["tooling", "auditability", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e8e31749d19358df370009266c32ebbc55dcbc5"} +{"id": "train_42249", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5767515a21755b2d0bb2d8bffc61f34336d7ad4"} +{"id": "train_42250", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e10aaa8f04fa2d533f1ffe5659a168655ec4110"} +{"id": "train_42251", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "security_gates", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "660e263b4a0fd14dd2aa39e1da2e0a75aca903ba"} +{"id": "train_42252", "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: 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": ["cost_latency_tradeoffs", "documentation", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7010014d311694a54d7cce8930a7e79f3e6099ae"} +{"id": "train_42253", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e71ec5f4abf3d0a59ba81fc4d441ef49de918e5"} +{"id": "train_42254", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fbd8245b1e28be0b6efe6958ac7f6a719f089d1"} +{"id": "train_42255", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcc7a222cbf2474c953e982f10974f00198c4473"} +{"id": "train_42256", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cce97c47af17e79656701c1d0636c846e0fbca6"} +{"id": "train_42257", "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: 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": ["auditability", "tests_are_truth", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70d810987bd4fe5aafc879f545fb3a2ee2379042"} +{"id": "train_42258", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b26c774a268f08d03b1c7aa72f4dc0da86ed668"} +{"id": "train_42259", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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": ["auditability", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7f474a822eb353689c0308de352ecfeaac23018"} +{"id": "train_42260", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "auditability", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b611c34d7509f23f6e68ad8548ad5b9c159da334"} +{"id": "train_42261", "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: 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", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "986fa607ac2986c771a312615e1ee3a734745917"} +{"id": "train_42262", "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": ["documentation", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afa9982da51ccad72c0aeee14dfd0e0b921ec562"} +{"id": "train_42263", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "tooling", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c456bde202227d52099ab4b8b3a652d762f8cec"} +{"id": "train_42264", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "documentation", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cb1b7a32989f7d61a3bb84c2ee5a0c6a33f270a"} +{"id": "train_42265", "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: 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", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b52b8e16ea9548a061c0671c3d021141ac52d5ea"} +{"id": "train_42266", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fae98ff39059acf44bc156306da2d70ac59310c"} +{"id": "train_42267", "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: 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": ["governance", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5ee251a80eba53c9a22efe6dddaa400cddc93e2"} +{"id": "train_42268", "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: 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": ["documentation", "governance", "tooling", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9472e173e180705b19d2d6b9f25fdfcc4d821b0e"} +{"id": "train_42269", "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: 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": ["evaluation_metrics", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f56becf30666193531799d9e090444ab3f3499d1"} +{"id": "train_42270", "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: 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": ["repo_scale_reasoning", "reproducibility", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f67f8c0969aaf95fe750dc4401855deba215fd64"} +{"id": "train_42271", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bef0bf1b2dc941a52195b3ad7f3d1db4b11bd5a"} +{"id": "train_42272", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["repo_scale_reasoning", "tests_are_truth", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e629f46f4ec2b4026b197b15333e68c6fbef4b1"} +{"id": "train_42273", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a98957c3085b218d4f57f720d26e9bdea0e606f9"} +{"id": "train_42274", "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": ["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", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b174c4706564962926a819c5f8e3df2cc1a3c64f"} +{"id": "train_42275", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c48ad8a70e5e6fac32ef73c234a002f2fa85e0e6"} +{"id": "train_42276", "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: 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", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "848e51e88ccfa2f1e1bf93e420278c03d59177eb"} +{"id": "train_42277", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbe7c45eefe36ad61f857cd08bb574394c1724c0"} +{"id": "train_42278", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecff726db656e92304e7099ac8205a982a345f32"} +{"id": "train_42279", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4001de3bd5886beb6542bc04a851e909331c4f30"} +{"id": "train_42280", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86e17ec31babe7bec15c889ed355f8f8fdf5fb2e"} +{"id": "train_42281", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "auditability", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cb4cf40d89c3276365dfdcc7ceb0689cf7beed6"} +{"id": "train_42282", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4ff8b8a97f8fcd08529bfd5474f810b7b33d3f3"} +{"id": "train_42283", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tooling", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a81351edb899ed6fd02359f2d344f2b57fdffbf"} +{"id": "train_42284", "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", "cost_latency_tradeoffs", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e743e5e89f4672b4fc2ef1fed93615e3000d4461"} +{"id": "train_42285", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6a5e2e9fbc16318e96f561de4775083d089f20d"} +{"id": "train_42286", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dec590b1fc2b290070e9dd16d45fd8998480f3f9"} +{"id": "train_42287", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac675704c47a66136e9ef7e4453ca0009c7dd57f"} +{"id": "train_42288", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d74931321744b25fecff8b48cb56f4857b3620a"} +{"id": "train_42289", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "577230705a404d0a3d095e6941524c463c377f1d"} +{"id": "train_42290", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97282068b17440043416189e4ef7dcedab40c3e5"} +{"id": "train_42291", "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": ["security_gates", "governance", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5d04c85560a48c645ef4e4fdc3681a90401c3ed"} +{"id": "train_42292", "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: 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": ["governance", "ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7201f24c603e1f711de8f81e9383e96ca542d9b0"} +{"id": "train_42293", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc906cd3339fc319dcde11794ff6c01882540440"} +{"id": "train_42294", "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": ["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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76a46615dcc12079f79e617b651d1a619c0611a4"} +{"id": "train_42295", "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": ["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", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2987fcfd08395894c82948be8f8310e60a87cbb3"} +{"id": "train_42296", "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": ["evaluation_metrics", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1b4664aeae25e2211874d7252b4a000f5a8eca5"} +{"id": "train_42297", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a14ab56f597856139cdeed989e7a7d5b4dbf6e40"} +{"id": "train_42298", "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": ["ci_integration", "security_gates", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82e35e4d43cb17d0d973c363d9cc55202a50b963"} +{"id": "train_42299", "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: 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": ["reproducibility", "auditability", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b00be2836bfd6c1b88dae9e5b4d8bc0b8983f7a1"} +{"id": "train_42300", "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: 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": ["governance", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5533a7d0f08e59927089cf6bad3b3d41a8ea8d4"} +{"id": "train_42301", "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": ["reproducibility", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cbdbb46984c0489627834d9b17d0adfefb30900"} +{"id": "train_42302", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a422a74f9bb27050bbc25b5a36f10b3081db527"} +{"id": "train_42303", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45d6e0b3ee4641465c2b95a7645f3aa239447aca"} +{"id": "train_42304", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0dca8c83554087393cff8093738f04ed09ab5a6"} +{"id": "train_42305", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5cb65f56a3925d4416b6513d4e80d12ef1edf87"} +{"id": "train_42306", "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: 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", "governance", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "699f7fb3b77e989c6d3fb751243333bc52cff1e9"} +{"id": "train_42307", "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": ["tests_are_truth", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1137d3adce98b56bee53c9205d61849703415b1e"} +{"id": "train_42308", "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: 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": ["ci_integration", "tooling", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23424c81e0d8c4da2c5d6f0c3c0be2f9fc39590e"} +{"id": "train_42309", "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": ["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", "security_gates", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30efd5b79bd76b2957ab26639b2147e36badd8ca"} +{"id": "train_42310", "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: 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", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "595f941516057c0ccaba1feaae84cdc6f3f65819"} +{"id": "train_42311", "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: 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", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37bff38f43bfe58403adce4f8d2a6cda06b7c1e2"} +{"id": "train_42312", "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: 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": ["cost_latency_tradeoffs", "governance", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "550b03cc5f8c6cbd954d07833013a144794d1f89"} +{"id": "train_42313", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1ad92aae4ef4d7585829795917a4f2ad9b5be9c"} +{"id": "train_42314", "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: 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": ["security_gates", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "007a13c606ccccf283138adb87385e2852c9abf4"} +{"id": "train_42315", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc85d8b1b9c9b7f872cf250a223bfe5fa03c87b5"} +{"id": "train_42316", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["governance", "repo_scale_reasoning", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71a417e10382fb493c8242ba152889f838662a27"} +{"id": "train_42317", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26b2466e0fcb42028a5c335594a8efc870b62498"} +{"id": "train_42318", "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": ["governance", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25e222563278a6732c7a1333c6186cbbb79beb91"} +{"id": "train_42319", "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: 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", "tooling", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "769a7c43a2b3aa59ca3c5155dd915d6bd7c7484d"} +{"id": "train_42320", "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: 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", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "683efe8824345fef79117a74c1c588e82877e708"} +{"id": "train_42321", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf943dbedc687a95cd214f7411c0ee3b19404ef1"} +{"id": "train_42322", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ac2084e33da878e6007536c59ca18596ef88467"} +{"id": "train_42323", "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: 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", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e3b322bdd956cb4ba9f535a243d549355104c6e"} +{"id": "train_42324", "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: 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", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54aa2fb0887910f830308a75c9e15f0c77d556a5"} +{"id": "train_42325", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25e0ded74f141cd6cc2d8295766d261ba25aeae7"} +{"id": "train_42326", "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": ["security_gates", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b126a8d0418210478d58df6b8c2add0c0271aae3"} +{"id": "train_42327", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["cost_latency_tradeoffs", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29a841c02be961e67a9f5206ea31a26f20b71000"} +{"id": "train_42328", "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: 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", "reproducibility", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee85eabcee6e4074a97bfb4395155cdbb0c57c65"} +{"id": "train_42329", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf943dbedc687a95cd214f7411c0ee3b19404ef1"} +{"id": "train_42330", "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: 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", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1de4b60f5a3236a66d2f9398ef70ecc9fc7dda06"} +{"id": "train_42331", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f15510cad5c70a5483de72ccb9f780645175b069"} +{"id": "train_42332", "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: 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": ["tests_are_truth", "reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd9b70c2f7da16c57c5fac38994d27238bc6912c"} +{"id": "train_42333", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff99bad3f45a21f23ff3fe2b7703bf9bd37065c5"} +{"id": "train_42334", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65763135e703a8562fb4d64b9feefe981515f267"} +{"id": "train_42335", "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": ["cost_latency_tradeoffs", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d33de789004b348a66d1e8a4070ded90fe841275"} +{"id": "train_42336", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef589e6a9d74704efe6bb12bb6a22c9b1d7cd7a8"} +{"id": "train_42337", "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: 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", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "331b209a41b72f07d40a6bf5bb5abd943ae59b1d"} +{"id": "train_42338", "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: 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", "governance", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9f4b67b5cf19d3f826b1cdb99bf06cbcf7f15f3"} +{"id": "train_42339", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38d20241845ae96d5b33f4e67637915dcaa97079"} +{"id": "train_42340", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f914b6966c8707f3de1c85d2ed1d67300f72b07"} +{"id": "train_42341", "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: 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": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80509a0fee1913b055471ba6485ff881e595e739"} +{"id": "train_42342", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "governance", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "956e8b4b27e4b2f148ff1aaba14d1ce204b05865"} +{"id": "train_42343", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b095922fbf5166761c30e1ff9c54ff72fbd33e71"} +{"id": "train_42344", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ceee9cf8dc2230f08a93b2142324d5065accc2b"} +{"id": "train_42345", "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: 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", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e587a47e73b001f05ce29bded2cf1b210f4d1b6"} +{"id": "train_42346", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "security_gates", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0faca18872110944b64f4b8034d36f4d31642d1"} +{"id": "train_42347", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aab25c26ad9f6b8e2f9b7b56c460fffc5c584d4e"} +{"id": "train_42348", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6fff7e70a55299b7c7ddacc04a1d646eb8038ce"} +{"id": "train_42349", "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": ["tests_are_truth", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34c25e5e463d9ba480c71c48a561179c8a13df8c"} +{"id": "train_42350", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "089c152d02c946e67c49f17a19c4f502f5d0523d"} +{"id": "train_42351", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be063aec92b2195379ff27910c5a5ec3e950f335"} +{"id": "train_42352", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["tooling", "cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d94e9d06d1e322d01c640377f2c6b0b7b84dcfb9"} +{"id": "train_42353", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ab836d846eb99d6909fade875af7911f43c0f85"} +{"id": "train_42354", "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": ["evaluation_metrics", "tooling", "governance", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cbfee780d03e5602e3a9410ded5da4806ecf88b"} +{"id": "train_42355", "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": ["cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61fce0b2cabf47c28c49ff20bcfb05e9e1947c1f"} +{"id": "train_42356", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "325a36381ff3c1e098324e0842501abad4fd8bb2"} +{"id": "train_42357", "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": ["evaluation_metrics", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75b728fb9eb9418cb096420b11dca5feed9b0d1a"} +{"id": "train_42358", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1dc6d774140d77e450406da63649ad269bdcf090"} +{"id": "train_42359", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cc87ae79683495170804956536e7c5b2752ade5"} +{"id": "train_42360", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea53333a2fa92842a6266982b7d57ecd26d913ba"} +{"id": "train_42361", "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": ["repo_scale_reasoning", "reproducibility", "governance", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a4d8531fc9ce51cbb7cc61f8f3b20098d1844e9"} +{"id": "train_42362", "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: 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": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "effe54370f24e8a826d6834ecae516ff9488ca80"} +{"id": "train_42363", "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: 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": ["evaluation_metrics", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d12d6fd6d204a1bb98ac1a89e6e8669edc6b33c"} +{"id": "train_42364", "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: 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", "tooling", "auditability", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0fa2aba56899e7b3fca9b08fb527e9fdd75594d"} +{"id": "train_42365", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a549ae98c42843884ce6717af286aef848e4d11"} +{"id": "train_42366", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d382fdfcc3b084546be8ec49a912524b21edc6b"} +{"id": "train_42367", "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": ["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", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c00157a1bd597860754197e31366c3f8e4243c50"} +{"id": "train_42368", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8943c336f8efefaa9fe9ce4399428a943d99bd14"} +{"id": "train_42369", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18c76d50b6cb8da07f3551b054a381ce3cea8fbe"} +{"id": "train_42370", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a738be0b34d1bcfb8102888bf942ed480e8439a"} +{"id": "train_42371", "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": ["reproducibility", "ci_integration", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "291ea89d0657822bdc52d2476fc1978c6059750f"} +{"id": "train_42372", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3c55836b5e7d8ad6c524a7350b534d6caf67df4"} +{"id": "train_42373", "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": ["evaluation_metrics", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7ea0415d92302841858ec39471a24920d245041"} +{"id": "train_42374", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a23756356691d1e738cf26e327ce76445227636"} +{"id": "train_42375", "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: 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": ["tests_are_truth", "reproducibility", "tooling", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "160f14673dd3353ede01c043efadc978c2f625fb"} +{"id": "train_42376", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92a04c10c7a1c2f551b2d118c2ff858e20353af1"} +{"id": "train_42377", "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: 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": ["evaluation_metrics", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10f5b24fe4e43d5823abf3f9d7dd1eaadbceffcb"} +{"id": "train_42378", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90362bf42f24e9e14f11340e69d99f5f48b3e7a5"} +{"id": "train_42379", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62f34e898772ae3d6b78f21edd3f0d7f0cee888a"} +{"id": "train_42380", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d7e995764e375ec45d043e84b8ab3ebfaa11649"} +{"id": "train_42381", "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": ["ci_integration", "reproducibility", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5798d5c5c3ae0b37846ac6df8bf5b987134626a0"} +{"id": "train_42382", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e54a5beb76970b3d40510463e4113336fd172b45"} +{"id": "train_42383", "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: 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": ["tests_are_truth", "evaluation_metrics", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82671206827793e1fd765d879f63e1eeb24a2dfc"} +{"id": "train_42384", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dec590b1fc2b290070e9dd16d45fd8998480f3f9"} +{"id": "train_42385", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56b404699850c3af148c0cea2c90818d9f404220"} +{"id": "train_42386", "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: 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", "reproducibility", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94367700160f03c2022e5755dd63cb3c159f45af"} +{"id": "train_42387", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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": ["evaluation_metrics", "security_gates", "auditability", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "196d9296e0c7efa9f40dac56bc0deb76bbd971eb"} +{"id": "train_42388", "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: 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": ["governance", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af4aed4c64f2c1e32bc45061081d643dd87a8b26"} +{"id": "train_42389", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aea24dec06e9bc8c861d65979e256bc7c28c2824"} +{"id": "train_42390", "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: 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": ["evaluation_metrics", "cost_latency_tradeoffs", "governance", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37d88806045aca688c0cdbf064574db3aeed447e"} +{"id": "train_42391", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5eb8054f2f1a48daef177a5e2bdc1810924c5501"} +{"id": "train_42392", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b21ae0aeba9b17fde5cace055dea2b09179c2ddf"} +{"id": "train_42393", "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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f09b499fa2456b41b7696a21d520caba0d1e363"} +{"id": "train_42394", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "726e1ccfb0a59e0722bebe2c03052150043fb12f"} +{"id": "train_42395", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b3bf32dd7de308b02d33916c00711fb2046644c"} +{"id": "train_42396", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f39df8ae7f593490ff6bab8e30fd1215f2e5d753"} +{"id": "train_42397", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "documentation", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1cbbaa5de2f0cbcc00af8016c1cd63e98fb5804"} +{"id": "train_42398", "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: 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": ["security_gates", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82165521e798a6c74ba1c6bba49acd35b5b95b2f"} +{"id": "train_42399", "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: 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": ["documentation", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "979894975c4616152dc814cca69a4df02ebb4c28"} +{"id": "train_42400", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["tooling", "evaluation_metrics", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47919346305490f4fa9f26359ded7aa4ab98e462"} +{"id": "train_42401", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7df8d565164079cc75d4baf34fb027d998147f78"} +{"id": "train_42402", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "security_gates", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5024414111403eeb8029d1928e83880a0df3faa5"} +{"id": "train_42403", "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": ["auditability", "cost_latency_tradeoffs", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3a7be1615f13557524072c9df6ac0eac07b9b12"} +{"id": "train_42404", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "documentation", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bed6bdb0c11a5ef93b23e7ac6c2cfb37222e0008"} +{"id": "train_42405", "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: 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": ["repo_scale_reasoning", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7647694813bf84062ee6a76e8fe3f5540a2b9c76"} +{"id": "train_42406", "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: 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": ["security_gates", "repo_scale_reasoning", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47d4ba9d988b3c976992f2688135257de31c56d3"} +{"id": "train_42407", "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": ["tooling", "documentation", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cd81a5b60b59ddc08cc99e2cd9778e8a9ea8442"} +{"id": "train_42408", "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: 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", "tests_are_truth", "ci_integration", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44e35fd43e70fe485de38ea906295df3a69ecce8"} +{"id": "train_42409", "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: 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", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ccf33ec741ed201764d23895d648f33fd2431c0"} +{"id": "train_42410", "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: 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", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e8284d0f2032a5c59b43ca45bcc5fe71ce2d512"} +{"id": "train_42411", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1ad77e2e9e682caccd67be6a8c0a7486f5b8a09"} +{"id": "train_42412", "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", "governance", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f56bed9b7ce6e49092da83a6570d63af613bd8b8"} +{"id": "train_42413", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f00a85d5180da91ccc97d7cab8b53f15987acb80"} +{"id": "train_42414", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca7102544c933fc0ab8ef6390b3a414726339d1d"} +{"id": "train_42415", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1afdc424f7dd2bdca668d782f87b7ab9c32d48ba"} +{"id": "train_42416", "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: 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": ["ci_integration", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7f197767d359f52edaf1d48f51ef1ac6171f704"} +{"id": "train_42417", "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: 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": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7010014d311694a54d7cce8930a7e79f3e6099ae"} +{"id": "train_42418", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "418c639c6ad717bbf0c1d191440c03fc8f473b1b"} +{"id": "train_42419", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea573220cff5494d14d33da02432bbac65eb8842"} +{"id": "train_42420", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33b525619799a54dca8e64ad4fa839350dec7d81"} +{"id": "train_42421", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c52003a0bd28dfbaea3687b6c92bec0693f12b40"} +{"id": "train_42422", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec698cc5d6afe1289af0473c52971ad2ddd96899"} +{"id": "train_42423", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d76d0e3bf459fe370808979489c31ea566255be"} +{"id": "train_42424", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c18a267192d51d4f7736c51db1e367798df4d8e"} +{"id": "train_42425", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f310d4f4bf247196a93b5e4e8c1e7242145256a"} +{"id": "train_42426", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "tooling", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2456e30cb5bbea480d6de2f98b389222e01350f"} +{"id": "train_42427", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac22cc1bdf3509bab3c3929fcd00d427d1b4fe4e"} +{"id": "train_42428", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b2c3afeb371ed5e30a836afe09b3e2cdf07c594"} +{"id": "train_42429", "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": ["tests_are_truth", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c49abc24076dce99d757a3b6b7f4d8b1fcd766c"} +{"id": "train_42430", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00e6a6c0eede6038c05876621b8b0daecc710d5f"} +{"id": "train_42431", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cb13ce8d15420f9eb018b0f4a5c70ade7a8b07d"} +{"id": "train_42432", "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: 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a59c6c72eb00fa0b7f2b9b37c404668e41638ffe"} +{"id": "train_42433", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5753f5f6dc68f08c1c1ca7a6e6901aa6f06df7b"} +{"id": "train_42434", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daf1afb7c87ac66e57bd3ce17df6a05c5c923a4d"} +{"id": "train_42435", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c3b543b060ee641ee1a56dfba874bbd8fae471e"} +{"id": "train_42436", "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": ["ci_integration", "governance", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e7ac289ee1cfb5830ae554724a17235da2ca5a2"} +{"id": "train_42437", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "tests_are_truth", "reproducibility", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "edf5a108d7e1e0158248e546639e154d7f5d55b3"} +{"id": "train_42438", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "080982ea5b1ebc0ff7c54c33d881461131c7d6b4"} +{"id": "train_42439", "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: 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", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "959a869a197c335eeb1a30e375918b2bc8ee452a"} +{"id": "train_42440", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fefa5c3f80141f9fe15cf2563a72af2581199ae"} +{"id": "train_42441", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b08fd811abf7b628571003c73c6ee0ac49bde3e"} +{"id": "train_42442", "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: 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": ["cost_latency_tradeoffs", "governance", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dda01b8097ec86e5aa2aaa0c5a43079168e94c21"} +{"id": "train_42443", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "032b8b6e2efe7d8313fc9c58670b29ca0b8fda1b"} +{"id": "train_42444", "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: 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": ["governance", "reproducibility", "tooling", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "725a1a0cd9d5cd4339e3194f70caeaa77a34f4b2"} +{"id": "train_42445", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e63c563aa75db0d4e93207ec32c87819810dac1e"} +{"id": "train_42446", "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": ["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", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82e35e4d43cb17d0d973c363d9cc55202a50b963"} +{"id": "train_42447", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3103404fdfa7847617da900b864a6d86145366d0"} +{"id": "train_42448", "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: 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": ["tooling", "auditability", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50ad995ae6acaf0657dc41118267226dddaa4cfa"} +{"id": "train_42449", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66d69fed7220905aeb7545ee1b05e6794588c9f8"} +{"id": "train_42450", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f1b0822b3ab47d89c5a64be3f28479c1ecd923c"} +{"id": "train_42451", "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: 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", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1e07f99e323faef5ab45485d4205974fc50b51d"} +{"id": "train_42452", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "security_gates", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f319def27e64737d01d39237f09bf41ae54813d6"} +{"id": "train_42453", "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: 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": ["auditability", "tooling", "documentation", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65708bebf748ca275fabb103aa632ca480830b5a"} +{"id": "train_42454", "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": ["auditability", "reproducibility", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "932a63bf8fcb16b1dd6163fd3e2a2f5241c3ba7c"} +{"id": "train_42455", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9356ae7905b35a1f7db92e751581b686a3f6936"} +{"id": "train_42456", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc735220db6fe5787d570ddd1d27136572876bf9"} +{"id": "train_42457", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c416f5f7885874eb847d327bfa9a9b0d8b5a803f"} +{"id": "train_42458", "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: 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", "tooling", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2477e8733776ce0c4d8aeec9a8e82c896eb1dbd6"} +{"id": "train_42459", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72016308604279e3516d2f2edc71dd5a82fc4bfa"} +{"id": "train_42460", "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: 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", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4aa0627b25054b52750d651fccbce655cf591a5b"} +{"id": "train_42461", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["security_gates", "cost_latency_tradeoffs", "ci_integration", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b66d5f6aa4118d387c9a4baea4681459471f00f"} +{"id": "train_42462", "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: 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": ["cost_latency_tradeoffs", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1554ddd6a87917fbdac944f808850442c06d39aa"} +{"id": "train_42463", "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": ["security_gates", "documentation", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e1325da28c8387c75123d690fac7c1f54601f6c"} +{"id": "train_42464", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["documentation", "security_gates", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea6b796b376c413588ebcf19fb1a17d36f0e9b8e"} +{"id": "train_42465", "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: 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", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fea2debfb4d82f9077bbe128ed0dc65712f2460f"} +{"id": "train_42466", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "governance", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a2899339895c2700638f0fa48e938cecf062da2"} +{"id": "train_42467", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca603e6d323a6f30a0f0ade6a535a812f4315d21"} +{"id": "train_42468", "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: 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": ["documentation", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b8c535d2afecf6071dd5e8ee9aa6ec63deeb29f"} +{"id": "train_42469", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71652407bf062f7e19c8fca844b18edd5f748831"} +{"id": "train_42470", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31cca24a8ae3b8589fa0946e170f28b47559d9e7"} +{"id": "train_42471", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48c16783e3f83634f05d43fc731201b7d7291ded"} +{"id": "train_42472", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["tests_are_truth", "security_gates", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74b871dd805cb0764f352f981a59c098cbfc3324"} +{"id": "train_42473", "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": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0970c907721c951b5ad61275587c557069bec019"} +{"id": "train_42474", "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: 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": ["tooling", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fea2debfb4d82f9077bbe128ed0dc65712f2460f"} +{"id": "train_42475", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29eef297b2793400707b763ef77bf2840f758311"} +{"id": "train_42476", "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: 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": ["security_gates", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13ca306b1003345b222a09d8e28392b70f3d123c"} +{"id": "train_42477", "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", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5712e5a6abda5fb9e4c877b0d6a723b8be7d1cee"} +{"id": "train_42478", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b51e1cff05cb2ff15a4d91fc59f927e277465c07"} +{"id": "train_42479", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f780144cfc67154a6d95f59b7d6b2cca53b75678"} +{"id": "train_42480", "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": ["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", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5f46ecd888c55cdae2636b7c10bfae204e01597"} +{"id": "train_42481", "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": ["documentation", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4932a49c7a45c8bf72d90ae171d6aa39673b68df"} +{"id": "train_42482", "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: 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", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11a8cd245787524c002c11e54c3df723d747eb7a"} +{"id": "train_42483", "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: 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", "ci_integration", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "241fbf8d89be04552a372c0c38bf8d7f32829020"} +{"id": "train_42484", "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: 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", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f3357e04defac47f5ced30fd6a8184e0ff6265d"} +{"id": "train_42485", "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": ["repo_scale_reasoning", "security_gates", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49215e6c7fb5a554bd1c072c802424182587d8d8"} +{"id": "train_42486", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2dc857ce28ab38dfa9b8c11201b7b4978f2d939"} +{"id": "train_42487", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1de4eb99959a5ee92c6ee32c2388dfe388f62a39"} +{"id": "train_42488", "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: 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", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d51b6c72869e9fafe4fc80762b8e1c079e5111f"} +{"id": "train_42489", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "reproducibility", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b29c55864ada0c4f40268727946382d867621620"} +{"id": "train_42490", "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: 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", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5655d6467b3c0f5cbcdda0c8cae2616572965299"} +{"id": "train_42491", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90b4d48959fb65baa9c4851ba3d066960f8c7653"} +{"id": "train_42492", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e215dd06e028166e10ee262911592afa8c61ff00"} +{"id": "train_42493", "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": ["evaluation_metrics", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5661dbc5a47e63f243ec8e9f4476925867f657d"} +{"id": "train_42494", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65203b78129a603510491bacc529f0776179b0e1"} +{"id": "train_42495", "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: 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": ["tooling", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ee57b681627ae496424a5edda145f5e39030a09"} +{"id": "train_42496", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ed7e24fbf054de36059c3e7fc06ae443fdb70b8"} +{"id": "train_42497", "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: 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", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c82c0835d8984587cd191e92624daf9753c7323"} +{"id": "train_42498", "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: 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", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "718e103d48919e91941db321175dce71d4c1ae64"} +{"id": "train_42499", "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: 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": ["evaluation_metrics", "ci_integration", "governance", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e85b8008fb78fa3ba0bd6a5431bcc5b6027c08a"} +{"id": "train_42500", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b893c3b8da8a81ae23d56b41cff55334bef6e99e"} +{"id": "train_42501", "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": ["reproducibility", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48e91fd54557a76e782766945ce6400965e5f833"} +{"id": "train_42502", "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: 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", "documentation", "tooling", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "300daae4965457447aa75b44a230d33514728ff5"} +{"id": "train_42503", "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: 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": ["auditability", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c79007190e5cd1cb2e013a1afdf69e06658ddd93"} +{"id": "train_42504", "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: 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": ["repo_scale_reasoning", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f4edb39f8d0007a1d99db0cc942ea490415112b"} +{"id": "train_42505", "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: 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": ["repo_scale_reasoning", "documentation", "security_gates", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9793ef3f74dd6c6c8e31a41122b9083f8bbeb9b"} +{"id": "train_42506", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe6f5db8150526c5a29fd5baaea0cd4d4b121156"} +{"id": "train_42507", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5318209d4d427ce284074928aeeeb7328701b8d"} +{"id": "train_42508", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "928001608bf666a2547c4baab7aa58813bec99f2"} +{"id": "train_42509", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85c72fe1b1c6653efae1f1ce713def4acf7a8461"} +{"id": "train_42510", "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", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "535b1f435d1b2109366c548bcdb295feca6dda95"} +{"id": "train_42511", "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: 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", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3794fb293f27239243741baaa0e66f6de2b516c5"} +{"id": "train_42512", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "documentation", "reproducibility", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "365356b0521319767702b3df054731e82ee7292b"} +{"id": "train_42513", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "repo_scale_reasoning", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e9a1e3f27f977d2c8c69b4de815bc30a84c3b46"} +{"id": "train_42514", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "tooling", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a265d862bb6f00b5e8e0d5208c6087785b0e12f"} +{"id": "train_42515", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["security_gates", "ci_integration", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af2b8c3eb6353ad82adb36b01da55cc52bf6c05d"} +{"id": "train_42516", "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: 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", "ci_integration", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8d3288252dd8c3efed808c43a61e75d345cf5d0"} +{"id": "train_42517", "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: 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", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c16b1537e3fff653656886edf5635f2f96f5c272"} +{"id": "train_42518", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8142fe97ad66a1e4056a07b7b405f94b76fa60fd"} +{"id": "train_42519", "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", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eee0353682455c824cab04b8e420414c84113b13"} +{"id": "train_42520", "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": ["documentation", "repo_scale_reasoning", "reproducibility", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76af5f70f701944b18b0c00c539b2c450f02fd5a"} +{"id": "train_42521", "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: 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": ["cost_latency_tradeoffs", "documentation", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19a21bf75370ea6ab04adb6b2ab2b6681eeadce8"} +{"id": "train_42522", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5cb65f56a3925d4416b6513d4e80d12ef1edf87"} +{"id": "train_42523", "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: 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": ["tooling", "documentation", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a41383309d501e394bfa67613cc1de414a345f0d"} +{"id": "train_42524", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bad73bb0f5691fef02438bdb90fa4ffc2ad02413"} +{"id": "train_42525", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3780dc897a9fe4ce30ff1ae642254492d04fa545"} +{"id": "train_42526", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1bdac09e6b460fb070108fa3ef530dcb3c93f39"} +{"id": "train_42527", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["repo_scale_reasoning", "tooling", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b857603fb7d8fc9ca1b71bdd45faf5e3bb30be43"} +{"id": "train_42528", "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: 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": ["governance", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28b28bf769bb8029ff42fe84a39d51e1a89a939d"} +{"id": "train_42529", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "817744671e204d1bcb4f5f2cc1108c42187777fa"} +{"id": "train_42530", "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: 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": ["security_gates", "ci_integration", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "047b51618d801132fbb91b342b8e2aa8b32ee748"} +{"id": "train_42531", "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": ["repo_scale_reasoning", "reproducibility", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f18ce8adf3f3a870f41706dc054399bcd6e89626"} +{"id": "train_42532", "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: 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": ["security_gates", "documentation", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd378791a7e9777151f27ccf260b29c330e4558c"} +{"id": "train_42533", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bbb9f54329a0d9eda29395a850ee6d052b13edb"} +{"id": "train_42534", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1024fa3569426561e444de15811b9d0ca258f0b6"} +{"id": "train_42535", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1901a39a900a2dfc0fe04ec960eaba463d5c4367"} +{"id": "train_42536", "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": ["evaluation_metrics", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59fefff0062fec7532a5f191e8051c57e5cf900f"} +{"id": "train_42537", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["documentation", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cacb495082b1fc03b4f96631bb5140486477fc5"} +{"id": "train_42538", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1e34154400abe6f945c7361863c72031ad2c033"} +{"id": "train_42539", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35f6a8012bf57f352b3fd75d3997b1eecb204ceb"} +{"id": "train_42540", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1654a04a9a4508448d887db9381a2fe3dfe29a47"} +{"id": "train_42541", "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": ["reproducibility", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d58a125cedfaee93464607b768803c96aeba1e2c"} +{"id": "train_42542", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31dd8f75c97a68879036aff68c5e94481a2b0486"} +{"id": "train_42543", "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: 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": ["reproducibility", "ci_integration", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac73df952c088e4ee7b542f6b70bc619dcca3076"} +{"id": "train_42544", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff75e54caa1093bae59bca6c9a45745e990611fd"} +{"id": "train_42545", "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: 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", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d695d3816be2d2e8de0f388405103a96608141ad"} +{"id": "train_42546", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb201427cbeb6237ef6c2a688c951e5e9e35a54a"} +{"id": "train_42547", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4dfacf8c30d7336da7f3a01593745f5dbad5ab15"} +{"id": "train_42548", "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: 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": ["reproducibility", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccdd893ccaa1a46449f0b8fdf3af7ba276127489"} +{"id": "train_42549", "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": ["auditability", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "725f94befe7296127ac2cf81472451faa6f07212"} +{"id": "train_42550", "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": ["cost_latency_tradeoffs", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "32723a0ca2376f8e66d890b2d1cb75430a7558bf"} +{"id": "train_42551", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3525387bdc8a4a95ed8d61a71c943e52c62b96c0"} +{"id": "train_42552", "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": ["reproducibility", "documentation", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "357289dedfa6831c79c6292d23e8bae32d210085"} +{"id": "train_42553", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "834d95543024ff4c0f73b1e2ebc6fd40b8d9d60f"} +{"id": "train_42554", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "governance", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68424145e971815920b2fcfaf26d1b4d8785f642"} +{"id": "train_42555", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b090e561f9b6ff8a7ab9c2a35e335d7e211d2350"} +{"id": "train_42556", "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": ["security_gates", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9365f13e7de68b932a101a20a700934bd4b1b5b9"} +{"id": "train_42557", "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: 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", "tests_are_truth", "auditability", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52c8079e530a1246eb4e1405804625110ff991d9"} +{"id": "train_42558", "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: 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", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88e4dca7fe73a7cd0a37311a65765ce1c24a8b7a"} +{"id": "train_42559", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["ci_integration", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "878c177b2628e217737d468f28f18273ae9c9d07"} +{"id": "train_42560", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15d222a9a167b1b4d13d5ecbe35bf269eb906582"} +{"id": "train_42561", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f49e1649409389867b49ceee2ab76774d5dca873"} +{"id": "train_42562", "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: 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": ["repo_scale_reasoning", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cd68bcae2c5472cd343ee3e54ba9e99a4528d9f"} +{"id": "train_42563", "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: 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": ["documentation", "governance", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "870940ef96d3a3d726323dd757ee08d744a20acd"} +{"id": "train_42564", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tooling", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1caf334478234684f9161ee2a3d6516b6962d31b"} +{"id": "train_42565", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "security_gates", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "693f05be8dbc2ca492ff44f3a4992d72614b12ef"} +{"id": "train_42566", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["security_gates", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac523aa8af9cd5e753cac9b4ae5180128a5bb647"} +{"id": "train_42567", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "857a2c1aa7f2c117094768bc64b12dbb25fdd189"} +{"id": "train_42568", "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": ["documentation", "evaluation_metrics", "tooling", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "539d30f19f9f9f3c61ce6f9f3c0be1aa86c21b11"} +{"id": "train_42569", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "921309d5ff62de57b4f0e26fd56caddd030a24b2"} +{"id": "train_42570", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["evaluation_metrics", "reproducibility", "tooling", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f52e68d77a996c568f64dacb3cc71df0e58f3fb"} +{"id": "train_42571", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "evaluation_metrics", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "702463f857eb726877895d6a50249655d3b37007"} +{"id": "train_42572", "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": ["security_gates", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e281a8308799d381215fa2a733e7c1810dd9b67d"} +{"id": "train_42573", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad981a4919ebe1c9e9cf20d96021f979c4a20464"} +{"id": "train_42574", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6775dafb5136ebf7b3a1babbbe894954c9ea5fd9"} +{"id": "train_42575", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "ci_integration", "governance", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cb13ce8d15420f9eb018b0f4a5c70ade7a8b07d"} +{"id": "train_42576", "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: 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", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "936781168d13d124ddb0f30ee1f31198fb52f3c4"} +{"id": "train_42577", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6a8087eaacd92f60ec9dd01fe9c8c1503cfd0c7"} +{"id": "train_42578", "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": ["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", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0e216080364cab431a47b77a28bc73f577da022"} +{"id": "train_42579", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d66a93dd98055fd72ea76efddab2f7f4761a8f4"} +{"id": "train_42580", "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: 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": ["cost_latency_tradeoffs", "ci_integration", "security_gates", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15cace6f7558d4b29a8a0abd21276f41ab3dc5a0"} +{"id": "train_42581", "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: 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", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5afe4a0edef47555a07359f9cb562052a6c12736"} +{"id": "train_42582", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f397b7b427c80455fc2cb97d5849b24ba14396d3"} +{"id": "train_42583", "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: 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", "cost_latency_tradeoffs", "documentation", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65708bebf748ca275fabb103aa632ca480830b5a"} +{"id": "train_42584", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "463ef235f949682101474b1f40ab45f152a44e66"} +{"id": "train_42585", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25af732c27585847f710e2285254cfa565b4efbc"} +{"id": "train_42586", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bf3dd125247e36e7d06813cdd0004153083779e"} +{"id": "train_42587", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "documentation", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9166a02f06ff96b55a66885a4cfcbd0cba445ae7"} +{"id": "train_42588", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4aef6afcd8395a42d22e72f46af1aac04e88bc0f"} +{"id": "train_42589", "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": ["governance", "cost_latency_tradeoffs", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fa63c37d1b79399c4b91203efeb6872b08d17b1"} +{"id": "train_42590", "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: 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", "auditability", "tooling", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1031c6575a124bdd56816d0ee214f9761dce694f"} +{"id": "train_42591", "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: 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", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "085642e87be9a435f1087afbfc26c1c0011b5df5"} +{"id": "train_42592", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60d2b4438119fde49252621785258e271f72c236"} +{"id": "train_42593", "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: 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", "security_gates", "tooling", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c6f651767365db0bda4887959c5bea19d047f13"} +{"id": "train_42594", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fa2dc911447fd9154ffa13a6ad0496fbbaf1e15"} +{"id": "train_42595", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac765edb38930627533bc365a641f85cdc113606"} +{"id": "train_42596", "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": ["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": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afb3e4987f840517385ea0a1ced148400128c0fb"} +{"id": "train_42597", "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: 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": ["ci_integration", "reproducibility", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53c45b63380ec4f50d8fd29c9d50753a573edf37"} +{"id": "train_42598", "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: 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": ["ci_integration", "cost_latency_tradeoffs", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24f17b79d1ba0d0518c0d90e13a92c76eaeec55c"} +{"id": "train_42599", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5d52d866ad6ce6b860db2d16286a351c2fb8641"} +{"id": "train_42600", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["tooling", "security_gates", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d037878618ccca634011752ef0e33d75219ed1a"} +{"id": "train_42601", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7b0d99b541592fce004e4bc2cfee5a93d7f12e4"} +{"id": "train_42602", "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: 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", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b38a5ae8ee82515bb9de43661dd57f7727c5388"} +{"id": "train_42603", "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: 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", "documentation", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8f2edf467479e2ceded0db42aa849be4e776d4b"} +{"id": "train_42604", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6343f9e397f886063fc03c55c5cc48ceff496cbd"} +{"id": "train_42605", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9692da5ea434dfbe4ad7ebfcaf62589345dc086d"} +{"id": "train_42606", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "documentation", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bef596fc5405740086f0b45f3457bcef5eb0802a"} +{"id": "train_42607", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e69440fa48c6c7445997b94f18b591d88407a3ff"} +{"id": "train_42608", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0264ddb058285b24c9323ac7f2e4f28a1d00e6f9"} +{"id": "train_42609", "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": ["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", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d20bfdc1ab4014700305085cdadab2ae873d556"} +{"id": "train_42610", "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: 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", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c75d41f9117a12708f3b0f1f273cf54bd86a7ee"} +{"id": "train_42611", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07e0897f20704d5b591e94e4b67b201449af1dac"} +{"id": "train_42612", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "817bb89b0f21db34a0ce45aa84ef59b48b2f6c21"} +{"id": "train_42613", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c86954a449adfcc2e237bc607d4b048867cc04f"} +{"id": "train_42614", "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: 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", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6db7b5f43c53f464469a93b8aca0263fbc2e2657"} +{"id": "train_42615", "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: 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", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa8a38d4931585b8d6f449bc53293c2e23ad64f6"} +{"id": "train_42616", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08c3a1332ee40457eb6360d0efdd662b05f1b2e3"} +{"id": "train_42617", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3443eba572a6224fab2cde9e1043737864969aac"} +{"id": "train_42618", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf0bd9e3f8e6e3455a061061cc16f0b3ad95e986"} +{"id": "train_42619", "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", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b3d56992a2df4e7d9100fc037f8e471786bdbca"} +{"id": "train_42620", "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: 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": ["documentation", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e622f47194e0036d9be983607368c2707edc872d"} +{"id": "train_42621", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9750e36bb0a1c48ea6b52e72668fcb95a57e38e3"} +{"id": "train_42622", "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": ["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", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a6b83fac7e9ad52a4d97a6f53f082ec7301c9c8"} +{"id": "train_42623", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07d12c0af401edbfe79351ced0828e93a090f27a"} +{"id": "train_42624", "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": ["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", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dede9eba7d2283e8eb936d62b3fb96733cf8a97b"} +{"id": "train_42625", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e311bcc1622fa32e1fa6acb1d8f0f15545648f76"} +{"id": "train_42626", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f51235cffb7720179665c886e07cbf1c9eb1c7fc"} +{"id": "train_42627", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d66a93dd98055fd72ea76efddab2f7f4761a8f4"} +{"id": "train_42628", "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: 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": ["auditability", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44c6dca66302d5dcebff4b585cbed5009bf50440"} +{"id": "train_42629", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "982d600761fb5b1f3900781d3d63c06e62cd4e71"} +{"id": "train_42630", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "132369fe9450ec4cc8cfacb4c130ee74e5c8768e"} +{"id": "train_42631", "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: 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", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f67e9790428c59c13fbe258e19a9ad7a33ae7ee"} +{"id": "train_42632", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c5597dd5d9899b7d4e319e354e0296a47d3c7cd"} +{"id": "train_42633", "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: 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", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38a8070ff1ab065240c2411c7923b523bd2d7f2a"} +{"id": "train_42634", "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": ["security_gates", "tests_are_truth", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f63f270b2218f2a8760c10d9f35b04511ee1c016"} +{"id": "train_42635", "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": ["repo_scale_reasoning", "documentation", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c911c35f6adb480509adab50926a6f35964e35e0"} +{"id": "train_42636", "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: 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", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d57c209086fe0555d3042dc1f26a58a4ad91ee49"} +{"id": "train_42637", "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": ["cost_latency_tradeoffs", "tooling", "governance", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b111f5707d3553afe68d04583d122f01f738da1"} +{"id": "train_42638", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["security_gates", "tests_are_truth", "documentation", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0013f4f6066ccfa0e114b4e1dc5d22f54caa0d3d"} +{"id": "train_42639", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bbab08a6958b7b73fe9bc1090f6610b9d6b3b41"} +{"id": "train_42640", "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: 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", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36c7bd3a87ed2ad6458feb3394dc99220c52db27"} +{"id": "train_42641", "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": ["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", "documentation", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5c347dedaf7a3a1030b58107f4da2a4f8a19360"} +{"id": "train_42642", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "931322d5c22cdbdb5db412d784b6317a942fec89"} +{"id": "train_42643", "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: 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": ["tooling", "security_gates", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1a9f022bf3b8124218692441deaa3b832e1194c"} +{"id": "train_42644", "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": ["cost_latency_tradeoffs", "auditability", "documentation", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a97488409a3194c4ec682209301122ee330d1903"} +{"id": "train_42645", "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: 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": ["reproducibility", "evaluation_metrics", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8c3ee8e052abe1635f0ae59db99c6073ff08abf"} +{"id": "train_42646", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["repo_scale_reasoning", "documentation", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ce6641fe892c3f9b874c5e34151eaff4861b49a"} +{"id": "train_42647", "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": ["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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c92dac616e663a45d891970641ca850743411798"} +{"id": "train_42648", "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: 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": ["documentation", "auditability", "tooling", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2364e9fe8c825937a938440e853ccac7c42fb2b6"} +{"id": "train_42649", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "tooling", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dae18790b7006890d81afd26cac4c0d1dd85e5c"} +{"id": "train_42650", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "747e4b588811f5fcd6a9f09a3a20e5a7412b0534"} +{"id": "train_42651", "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": ["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", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e1b00da5daa5143cf9ed967fe3abee5bd4aa520"} +{"id": "train_42652", "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: 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": ["ci_integration", "reproducibility", "documentation", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a26a806aa0163ea0e970bbb07b6eaf2812c8e605"} +{"id": "train_42653", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "documentation", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1aab9a3ff0e08e0f05fa4b4ff989fc674f3cbc6c"} +{"id": "train_42654", "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": ["ci_integration", "security_gates", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5712e5a6abda5fb9e4c877b0d6a723b8be7d1cee"} +{"id": "train_42655", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93ffdce60c745ea9124d6fd022ffa381b1fdf2ae"} +{"id": "train_42656", "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: 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", "documentation", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08021b03d2768e03e3f4a67b9abdac50997e87cd"} +{"id": "train_42657", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "documentation", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4ae3317f4b12c9082921bdd9e73b639435f84d3"} +{"id": "train_42658", "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: 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": ["tooling", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7098ec03b17336245a906761e98ead0288ce69e3"} +{"id": "train_42659", "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": ["auditability", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef16b6adefc251cfb38b40e8adca697440988243"} +{"id": "train_42660", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe5375551eb42c69523d48b79406df3adef15e85"} +{"id": "train_42661", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15cd8b88eb9bbaa3c88bb28c9c7a1fba630aa327"} +{"id": "train_42662", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a2b0e5deb1645b93165858a30f0a426cc4bef21"} +{"id": "train_42663", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["tooling", "governance", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e899f432848221a42343c9e48e3ed4c77942012f"} +{"id": "train_42664", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e9c29839dfd37aa3678d4a67ab5116f4a53a2e9"} +{"id": "train_42665", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ae41b2f0d7eba7aa928a819bcd56f97dac54c64"} +{"id": "train_42666", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1eb59525d087f4a83ef1cd6a1d4d1a49ec0d631b"} +{"id": "train_42667", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "385d309ddd322c74a7802a28329bbad05c43a855"} +{"id": "train_42668", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01370dfcbb2f5267610aad90a9dd8f22a8dedffc"} +{"id": "train_42669", "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": ["cost_latency_tradeoffs", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "044f02f93d16e8c7871f59e0dcf17a80bd76b418"} +{"id": "train_42670", "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": ["security_gates", "documentation", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53d847bf7cb7b09b1d3a6596988f54eb84266e49"} +{"id": "train_42671", "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: 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", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca091d847ff755f29a45035d5a8cec55155c6d92"} +{"id": "train_42672", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "security_gates", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21ff77c7a56afaac255c38fd37f41f9ac8e17310"} +{"id": "train_42673", "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": ["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", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe713ab07e7387753db0ff0948c87090e58e154a"} +{"id": "train_42674", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89f967ab9082be207080bb254ffb40b412bb9cc7"} +{"id": "train_42675", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bcb79ed847af7a3bd6b24f863db794273e62dca"} +{"id": "train_42676", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "documentation", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "906c11c8733227f060d0a9193a674688b80a5736"} +{"id": "train_42677", "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: 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": ["documentation", "reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3187fa86c18e7144165b1990338ab04f32184f00"} +{"id": "train_42678", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["repo_scale_reasoning", "auditability", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01ba3baa9c6208551e0902d4a2d107cd5b43ea42"} +{"id": "train_42679", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88715535926a060fce1c07477f1233312456ce13"} +{"id": "train_42680", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f27d9bdfa6b25703a863f21a90ba21ec9f24fdcd"} +{"id": "train_42681", "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: 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": ["security_gates", "reproducibility", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "924c0507fefd78e67bbeb637fe30641f2901735e"} +{"id": "train_42682", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d88454134419e7bdf33789dbc5a9e0c3bdce035c"} +{"id": "train_42683", "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": ["cost_latency_tradeoffs", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb2a9301b8ac14b50ea16fbf2a0837c6f65998b9"} +{"id": "train_42684", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "500eaac51a4bda61540061e512dc7b675439df6c"} +{"id": "train_42685", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29e082738a49ce02e0fa1a123fa75329fe28b5d1"} +{"id": "train_42686", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["tooling", "auditability", "documentation", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60e47da7314d440dd3510bea0569c1a91fee492d"} +{"id": "train_42687", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a1f7ef37a1122b6582deddc6247a14a4136b47d"} +{"id": "train_42688", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9419c2722707b711c5c3ea56d4c7c3ed7ad27b2f"} +{"id": "train_42689", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc361847b4b304b67e51d8c68b8b719ac1b24bed"} +{"id": "train_42690", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d5bff36966c74b0c520f3f79054e4eef23e0766"} +{"id": "train_42691", "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: 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", "repo_scale_reasoning", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2ac6cf5fd468bac88f7ff0f5ca97ab502cd1f4d"} +{"id": "train_42692", "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: 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", "reproducibility", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a16017f05c88a069f792e9610aa72282529189b"} +{"id": "train_42693", "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: 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", "governance", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0acf90a6e78e2d052b707034f451dbed77b4d863"} +{"id": "train_42694", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "353a7540b7e0761baa39a29e11baf300fc6f732f"} +{"id": "train_42695", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06d22ac1448f4499d0cdf6bd91f55df5bad2cc51"} +{"id": "train_42696", "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: 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": ["governance", "tests_are_truth", "tooling", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2290fd207576056c24f37dc418658a32a10ba894"} +{"id": "train_42697", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1b0391d4849b7ea77be1e59a2930a9fd0980de8"} +{"id": "train_42698", "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: 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", "governance", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9619f2058850769d56bb577165f6b347e63f89ed"} +{"id": "train_42699", "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: 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", "governance", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cb6ad1b86237674a7c56d0fc3079649a0438ca9"} +{"id": "train_42700", "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": ["documentation", "auditability", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d06dba29cc434110560eb6549ec834777af299af"} +{"id": "train_42701", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "372a594efe0dc214730492ffe4a5caf6828c19d3"} +{"id": "train_42702", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4856d236b6d7e059bcb005acd5d4279e63d1d1cb"} +{"id": "train_42703", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4bfccc91748f6ae45cf0b3ec8c05c611090affae"} +{"id": "train_42704", "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: 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", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68e0120c91c1fb45a825bbe1df050e69b66434dc"} +{"id": "train_42705", "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: 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": ["tests_are_truth", "governance", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f310a8892b8a0230837a4e51ba3f58ce36bda32"} +{"id": "train_42706", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d49cad0a57f0e15288fd2fa64a696181f58e9a6f"} +{"id": "train_42707", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7037aa64f69b49bf12c9e1d3ff53de350606cde2"} +{"id": "train_42708", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["reproducibility", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01faf62f87d444bda4114f42d0481284484a1f63"} +{"id": "train_42709", "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": ["cost_latency_tradeoffs", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d2a84eb702eb67c43252d2d94d7cb815db255da"} +{"id": "train_42710", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93b81b9fcfa5157e691f7968e6e59bc15064916d"} +{"id": "train_42711", "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": ["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", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfc3f20585ebf4da1cd779ae83594351a19f3196"} +{"id": "train_42712", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28805155a35d2859e80f2f976ac93a62a9cab644"} +{"id": "train_42713", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a6a95396557bff2e651a6535f7cbe99d76df003"} +{"id": "train_42714", "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": ["repo_scale_reasoning", "ci_integration", "tooling", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe9523f8281f043fe9c56172ab706ef3850ed93a"} +{"id": "train_42715", "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: 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": ["reproducibility", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6730354c0bf2bcf86e7a754675b4cee54552f6e6"} +{"id": "train_42716", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19254a810261c6039c4a6a5371cb9247197ccae3"} +{"id": "train_42717", "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: 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", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1741699007051183f91436d0a67298a84c53299e"} +{"id": "train_42718", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b17722a529ef199737bc0cadac3282fbe9e0994"} +{"id": "train_42719", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31bcd8cb4a5b67258b98b1b49311ad20dd402c7b"} +{"id": "train_42720", "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: 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": ["reproducibility", "documentation", "security_gates", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94fbdcaf0bdd668cb300f8cd615d298debdd4a22"} +{"id": "train_42721", "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: 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": ["governance", "tests_are_truth", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5a522f7c6023fd2b6b0ea99468da0ada00a510fa"} +{"id": "train_42722", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "reproducibility", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60069090cb0d4ad3f4368bed9929e4eaa7a1cba6"} +{"id": "train_42723", "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: 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", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "616df2346a1ca9afce9769a1ed79843940570476"} +{"id": "train_42724", "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: 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", "auditability", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "580798ea05b8db0020dd77c6764705eed90cd808"} +{"id": "train_42725", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aff4b0f4eda028aeb34057e29b2e2abb26d51e18"} +{"id": "train_42726", "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: 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": ["evaluation_metrics", "ci_integration", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80d4574d5555eec06d07e0a9c9ce0dedbbb8c1cb"} +{"id": "train_42727", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cbf5d3624f7df50b0dea73dd9d25d46d50b7320"} +{"id": "train_42728", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfb0ae90860c5a0abc4c5d526bdac730214cd09b"} +{"id": "train_42729", "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": ["cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84dc69c88b1b0d29131081ba7559b36fee02e3ef"} +{"id": "train_42730", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66a655ed1d62463a1b1d9c80ea2b2cd21625ee84"} +{"id": "train_42731", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d23731ada95401261bfe3eb77bf7a801471128ed"} +{"id": "train_42732", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e6c8d148825a2112f57b6361f510ebcc290817d"} +{"id": "train_42733", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a19095ce24a4610417aba3096103c4416fb0dfa3"} +{"id": "train_42734", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef791fc2022727e3b1c9996c5cc996b32e59c18c"} +{"id": "train_42735", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b0152fc867cb73ce7f0a370c4069462bcf7cbc2"} +{"id": "train_42736", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "926b754563f02daf27756e13be3e0f8e4440f6d2"} +{"id": "train_42737", "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: 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", "documentation", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8c9bb67e2348a4f6d6e90f58e369c77a75c2e10"} +{"id": "train_42738", "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", "repo_scale_reasoning", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "480aaea3325a2a5d379fb0ef2a58551f820febc2"} +{"id": "train_42739", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5cfa2e8a69192e6c1218a382ddbf44cb1a08bd35"} +{"id": "train_42740", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e1781451418649c29a98a9ce14062098a24a460"} +{"id": "train_42741", "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: 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": ["reproducibility", "security_gates", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7334b94009b65aa5fe19237bcc9a0e1a86024ac8"} +{"id": "train_42742", "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": ["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", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6dd0fb14faf7cb13eef1f64be75d65ea68e7db2a"} +{"id": "train_42743", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["documentation", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b13a800a63d2609fbe52c1a967a3781569ec2e4d"} +{"id": "train_42744", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a855822365133247a3384698fb945b8d8232c33d"} +{"id": "train_42745", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e34ff5bfcb2031d814c8cfed2fba9679fa8eb3b2"} +{"id": "train_42746", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33b525619799a54dca8e64ad4fa839350dec7d81"} +{"id": "train_42747", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8da6080d73f1970c545a5968b4f6663c3b09c8df"} +{"id": "train_42748", "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": ["reproducibility", "tooling", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b8a68184df6741f884e413fc7cac30635e64351"} +{"id": "train_42749", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53e8ce31948c96c40e37995b85c598b7ad4e8e13"} +{"id": "train_42750", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9000c790d6e57474a278a34155656003b1c15386"} +{"id": "train_42751", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcc7a222cbf2474c953e982f10974f00198c4473"} +{"id": "train_42752", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "security_gates", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56229da2f426577bdbab396ebdf05207b484b76b"} +{"id": "train_42753", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ef34b3852083ddd34fbf315fa03d75a86b4c629"} +{"id": "train_42754", "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": ["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", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "029446c0315b2adae4f68f5c672fbf95d53efe08"} +{"id": "train_42755", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ecb1e0e131200969c2de4654f133486153b7053"} +{"id": "train_42756", "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": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3110e3da9bda398463ffc64cfdad8a55351bc93d"} +{"id": "train_42757", "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: 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": ["cost_latency_tradeoffs", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dae6346f3423ece6be8c55454905db59369ed42f"} +{"id": "train_42758", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37328fbd81aede19e63948ed2c1b2b689c970509"} +{"id": "train_42759", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "reproducibility", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41c1c9781c9c55f0c5e57f8b105469ce45c9a347"} +{"id": "train_42760", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["security_gates", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "904dc64e581713666ac9707c731bbc749205adc1"} +{"id": "train_42761", "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: 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", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f4d355f678d19bacd2909a640d13c5e054db8e7"} +{"id": "train_42762", "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": ["tooling", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42ca21e67b76fb6d442af95ce756ad1f33bed0c8"} +{"id": "train_42763", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "625239e8acff96c025e92819308d08076780c14c"} +{"id": "train_42764", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62e13a982f1f7d6d08e12d529725ead9881ad31c"} +{"id": "train_42765", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65e9e75812d0d84f2130f3668e1a6e2607330951"} +{"id": "train_42766", "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: 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": ["auditability", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b8f85dae81fa224cd9e36fb26c9dbc3125c3040"} +{"id": "train_42767", "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: 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", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36d211c05ae340cd917d751c7c8603f4ace1f621"} +{"id": "train_42768", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67264f8498b9ccf7e532b2156f5400783a83cc04"} +{"id": "train_42769", "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": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2d65370b816d103ae59596e64a63b0ad380d632"} +{"id": "train_42770", "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: 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", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b99ab5e828f7499063fbe302b9038cfb8330f454"} +{"id": "train_42771", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "security_gates", "auditability", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aabc477eb2ca5c772d75b6593ac7b2b74f260b44"} +{"id": "train_42772", "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: 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": ["reproducibility", "governance", "documentation", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdd53b12b75380266d3ead91905ed6f1e3d18b4c"} +{"id": "train_42773", "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: 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", "governance", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "115eed8757156b2a2161ced46c8a18fe247f4b5f"} +{"id": "train_42774", "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": ["repo_scale_reasoning", "governance", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a832f1e119b4b5bf3b2028a5042e739a4852ed13"} +{"id": "train_42775", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "befe6fcd82a21ff487101d92df08b891d6d868ae"} +{"id": "train_42776", "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: 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": ["reproducibility", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a87daa32bd1f37fad71cf767e4a5e9d2c079400"} +{"id": "train_42777", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc52838f40ac6aaf5c4db9ed8b727aab6031d074"} +{"id": "train_42778", "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: 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": ["evaluation_metrics", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e62d760397ee036e0ff9e51f7b1bd1173924bd0"} +{"id": "train_42779", "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: 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": ["reproducibility", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4cee9a9d9bd766285eef29b91df9807b9dcc9b5"} +{"id": "train_42780", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["tooling", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1971cc4ee490740e637f9d65f2e0af2d9ecd006"} +{"id": "train_42781", "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: 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": ["cost_latency_tradeoffs", "evaluation_metrics", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f85335ab5fc955790e3dd944a5c28431af0434d"} +{"id": "train_42782", "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: 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", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c70897e144508fb77980fec13297ee795f6ae1d"} +{"id": "train_42783", "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: 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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5fd5f26bb205f3257f16f534d3f765b1d8fc072f"} +{"id": "train_42784", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7076693ac4530d82345cdfece7cd2c7ca7965c3"} +{"id": "train_42785", "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: 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", "documentation", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2aa0c93d623f387c7126a54a14057fa7e268b11"} +{"id": "train_42786", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cf65ef710c039054f5f12a2b2ca968b73d67845"} +{"id": "train_42787", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfaa959a6529f7137ba77d843584ec888a167815"} +{"id": "train_42788", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e4509c248e774e0ef5e0c3c6b88383dbf96eb3a"} +{"id": "train_42789", "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": ["cost_latency_tradeoffs", "documentation", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8d5cf67c9fa01709bbfd7caa14007267bee5c7b"} +{"id": "train_42790", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["documentation", "ci_integration", "governance", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1f8650f1e2233c761550335b2e3b9a04d3c6cc8"} +{"id": "train_42791", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56d68f00b629e6d8087b9968114f190060e2e437"} +{"id": "train_42792", "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: 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": ["documentation", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8212290f9ff89abd4342ce07cd00d7e1938440ec"} +{"id": "train_42793", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e95be472ffdd416c81a6b27e97a1cdf07232c25"} +{"id": "train_42794", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9c5dca6bc639927cb35b65aaa14ed292641cd10"} +{"id": "train_42795", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bda6e8476910b41921137a1fe277ac232aaf78b0"} +{"id": "train_42796", "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": ["tooling", "governance", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2f1809636967336bb09d729e6b684dc9265fdbd"} +{"id": "train_42797", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "documentation", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08ca2c0e9af78c8694ab6331eab52cda57d401c4"} +{"id": "train_42798", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "250d6e03e2f97a7f18b115dacfe54e93b6420cce"} +{"id": "train_42799", "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", "repo_scale_reasoning", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5798c3bef9cc2c131327ae6e1e870397aa439d58"} +{"id": "train_42800", "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": ["cost_latency_tradeoffs", "tooling", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9e23de698b1d2a4a3fc33ee3ffb84b634a7c83b"} +{"id": "train_42801", "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": ["governance", "auditability", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b44c09964d2e7c43c067f4754170e145f459dd5"} +{"id": "train_42802", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9894db08ff55327f4957157eb45d85e80823bdd"} +{"id": "train_42803", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc2caf6ee9a23d0b37dab141d8d473614432756a"} +{"id": "train_42804", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3878f7d409632fde556389cc9dceab523d0ee833"} +{"id": "train_42805", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9729e283a8930aa99b0ac5133ca6b78d6b0d7dd"} +{"id": "train_42806", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34626fa2bb670662f92380e852e909a0a264df78"} +{"id": "train_42807", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "security_gates", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "726e1ccfb0a59e0722bebe2c03052150043fb12f"} +{"id": "train_42808", "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: 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": ["tests_are_truth", "security_gates", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e8e0fa741bc59173a47f82c1af5dae7b8f1bf2c"} +{"id": "train_42809", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2e291e33a4babe4e8c9084c919f54d03b214228"} +{"id": "train_42810", "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: 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": ["cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ec6275b1f5b3ddd4e119ddc2570469d4db1481a"} +{"id": "train_42811", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f690eb62963fa5cd508bffe325d7d5392bbcf7f"} +{"id": "train_42812", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "724939db5db965610acc40faa44ca68a894eadb2"} +{"id": "train_42813", "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: 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": ["repo_scale_reasoning", "reproducibility", "documentation", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8da14a0007ee6aabeccf594ab69e800b338a2978"} +{"id": "train_42814", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7a1287ba1165f848578642e1ecb7264b54e8874"} +{"id": "train_42815", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fc6480bffe5ed2bb8afb6043fdd82de1609151e"} +{"id": "train_42816", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "738e59870d9cec94390676c2fc93e1a423b97622"} +{"id": "train_42817", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70b6510b7d1048a0a750c5e1c216f9a36d768a49"} +{"id": "train_42818", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c04485d9a2ec5b160cf3c586eff7d4e856d78a61"} +{"id": "train_42819", "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: 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": ["ci_integration", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ae10e58249135bde12c7784a1073cfe65090d9d"} +{"id": "train_42820", "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: 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": ["governance", "repo_scale_reasoning", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ffa2ffc2094e64265a69133a307479acf46dcee"} +{"id": "train_42821", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dec44cc5b7f03d6c332ef86bda30481b83ed56e0"} +{"id": "train_42822", "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": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58e0b81ec58196536575947ef1fad5ff8efc47c0"} +{"id": "train_42823", "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: 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", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "397d5aa576e418b231f03b25da002206203398df"} +{"id": "train_42824", "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": ["governance", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4bb26ca493bd48a313b02edc47c393fac6d0018"} +{"id": "train_42825", "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: 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", "ci_integration", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56e6b5ee9851c35d53292c03fc8065e6f906a57d"} +{"id": "train_42826", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "32a7a6588521854009aab85f007a17c3531d8368"} +{"id": "train_42827", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "governance", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7763094b3273de143e3d2abb0b9f891cd89ddfa9"} +{"id": "train_42828", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d530dff81a1195ab85a876b5d29939372cc686d1"} +{"id": "train_42829", "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: 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": ["security_gates", "cost_latency_tradeoffs", "governance", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "992fb59989bfb5401162e45b5ff327c521a590ea"} +{"id": "train_42830", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec093006cc408c3a9e378778de2c8fe8cb213fb5"} +{"id": "train_42831", "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: 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59f8d762fdbbe3a75629193c74438f5d5e784a70"} +{"id": "train_42832", "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: 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": ["repo_scale_reasoning", "documentation", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "124535f824ad81f48e045b67f918d703e6a1921f"} +{"id": "train_42833", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21849a7c820160f4ba460b6689e707683e6553f0"} +{"id": "train_42834", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3652099940251296d5f518991c8c963b2b61015"} +{"id": "train_42835", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "tests_are_truth", "ci_integration", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5f9f4c76a9d8af65b4075e58b78a9c979b4aa61"} +{"id": "train_42836", "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": ["cost_latency_tradeoffs", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf4e71d4f16493e21266dd9f46d67d46cb02e6ba"} +{"id": "train_42837", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["cost_latency_tradeoffs", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b0dd3d0c11f4bbe2da50ed323870408a6fe68a6"} +{"id": "train_42838", "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": ["ci_integration", "evaluation_metrics", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a4290c70fd8b825cbd390870e0dec1095c4eb99"} +{"id": "train_42839", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b317000f88d205ee1ede4718d94b1138e95a468b"} +{"id": "train_42840", "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: 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", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d53b582b3acb1ea6837e47afca2fb595ba58e52"} +{"id": "train_42841", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4fb9a380da262bdd0d8fce65f4e5629233ac6780"} +{"id": "train_42842", "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": ["tooling", "cost_latency_tradeoffs", "reproducibility", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28ffd172ccab05a12512e2a3bbf89b1c1eec71d2"} +{"id": "train_42843", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa79b3025505ea402db5623c1cd07a1cbd3e3977"} +{"id": "train_42844", "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: 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", "tests_are_truth", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24f4dd561ccaf911cbea7e5f6a5b684c8fb1ee94"} +{"id": "train_42845", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e88fe0416d75cb98847aaf20916e9abbdbf515f"} +{"id": "train_42846", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db4c388c50e1e91f700f5829c4021dbfbe7c2b79"} +{"id": "train_42847", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "982e166ca53a5adaf2d9cd63a667a1f1cc954cff"} +{"id": "train_42848", "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": ["repo_scale_reasoning", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f048f79c5f7cc1236f937c415b281c80a3668554"} +{"id": "train_42849", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe4b5c84cd9f7a45009d2df35bb47651cdec50f7"} +{"id": "train_42850", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e87b0d26735b7d2b32f8fdbce774fa91b376b0e2"} +{"id": "train_42851", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a9a1a03cd4dad653ae3639d07609abe87ffac3b"} +{"id": "train_42852", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "tooling", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5f9f4c76a9d8af65b4075e58b78a9c979b4aa61"} +{"id": "train_42853", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "166ccf50976c1f2c649a92ea71900bb9f32c259e"} +{"id": "train_42854", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "933ecae5220c9d9b1cbd992be6cb32b99ea509a4"} +{"id": "train_42855", "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: 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", "repo_scale_reasoning", "security_gates", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5d898966be5477e6be0b9b37edccadfc4dbdd94"} +{"id": "train_42856", "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: 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": ["reproducibility", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cd6d187054be336bb5de3374e30e78e9dfc80f7"} +{"id": "train_42857", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "reproducibility", "auditability", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "299f02e5ec0fbbdc94e82f9ac010748f9433c0f7"} +{"id": "train_42858", "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: 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": ["repo_scale_reasoning", "tooling", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f6bf9a5bf4c2a34f8fb2bf3e00cd3c168f08b9b"} +{"id": "train_42859", "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: 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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca8145fd6cb788e00b9de2ba7961db651aacb310"} +{"id": "train_42860", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4697a90e56b448553bb6cfadd5f136370decdae5"} +{"id": "train_42861", "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", "tooling", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96d4fdff6527549f2d823c7998186a1a3fbe11d0"} +{"id": "train_42862", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6eb1078d20d0f2e552228288e6c1366a9169a523"} +{"id": "train_42863", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4474e925f89f10b086c7e0f4fe4382f709a33da"} +{"id": "train_42864", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da7dbb2e4a1e69c7c1f1db3e2056fdc2603aeb4e"} +{"id": "train_42865", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "309130896e7d8691fef8a924634be5c4cec5d146"} +{"id": "train_42866", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d376ef0fb8b50840debf0af6167c1c5b06b8954e"} +{"id": "train_42867", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "124aba0eb5dad3810473b6fa37882142a9667f5d"} +{"id": "train_42868", "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: 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": ["ci_integration", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b4c91c7c7d726f9b807a4b5bc125d8ad394344d"} +{"id": "train_42869", "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: 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": ["security_gates", "documentation", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d91e7971dc03d84142e1e6abdbca7e8a9a1ebbbf"} +{"id": "train_42870", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d965c663512115dd22caeb4936f8e9096421611e"} +{"id": "train_42871", "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: 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", "cost_latency_tradeoffs", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "452a0561b856ff4c2b4f207fc8c8bd451f4092ed"} +{"id": "train_42872", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "405838758f38f1b8b4e6a81932a5514b316baa77"} +{"id": "train_42873", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "reproducibility", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f711b09e61504412f9c32ce18950264ab67b8cd3"} +{"id": "train_42874", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "reproducibility", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f9c095c419df31593a30943e677218b49119823"} +{"id": "train_42875", "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": ["tooling", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5e7cf0667b14475c356ef48dba69a91ad6ef061"} +{"id": "train_42876", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb4e32c2f88f73ed91a6835b2148f55c2126eb7b"} +{"id": "train_42877", "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": ["tooling", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bff923dd4f65887cf72856ba328e9b2df3e3af99"} +{"id": "train_42878", "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: 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": ["evaluation_metrics", "tests_are_truth", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22d0e1c78c2e6cacaa2ff8fd1e6a7bb4b8973d70"} +{"id": "train_42879", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1be04c48c8332cda3d12713721cdbb2ca3bb05bc"} +{"id": "train_42880", "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: 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", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "956f1da0160faf53f6dd70395caf201a9d72f137"} +{"id": "train_42881", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "721731e2b9160beb4f329626808cd4fc88afccd5"} +{"id": "train_42882", "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": ["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", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6f90a393e4caa8b28cef784bf28b3a2c71e54c6"} +{"id": "train_42883", "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: 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", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ab131190ef3ab4e8336ef8beb37ea4e4df6db58"} +{"id": "train_42884", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c2bac95d56e3569ff5535738d1022bff4a700b2"} +{"id": "train_42885", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "documentation", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25143162f230c6780081c708795f96ccd0da37b6"} +{"id": "train_42886", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d3ea98e925864bb10520d499f6d4f9c4f406bf5"} +{"id": "train_42887", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c29c84e0ef5abd3cb95dfb4a5512579d5e60b9c2"} +{"id": "train_42888", "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: 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", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "449e8aa7daffdcd92c4a729c9dc89643eab1cd0c"} +{"id": "train_42889", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "documentation", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ff00aa2ad098c4adf4cf1107399b6b70e7e7fc"} +{"id": "train_42890", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c18090f636de1b0dc9a011c9af4b407b50822ab"} +{"id": "train_42891", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f4075effe229aa27c83bbd0f74d6f0e0d904a28"} +{"id": "train_42892", "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: 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": ["documentation", "governance", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d8b4f463fe2d4a51d2ea0097b31d58ef37f004b"} +{"id": "train_42893", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da65526316801e470bba5adff4bc6767a04dff90"} +{"id": "train_42894", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e59645475f3ed9608d6fe900b90dc0bfbfefd90"} +{"id": "train_42895", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc89cfd636dfc20943ac5e2dea2c2d6c04919f59"} +{"id": "train_42896", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92788372a172f6ed02628a5cdf589cf477046908"} +{"id": "train_42897", "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: 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", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6137da2e68581f48125c01cd8fbfbfda88752f2"} +{"id": "train_42898", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["repo_scale_reasoning", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7d2f594a0e5d21a5d34f9d1d363921bc228f72e"} +{"id": "train_42899", "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: 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": ["security_gates", "ci_integration", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51502071c280db9df313f35300c23d58f358c102"} +{"id": "train_42900", "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": ["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", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99d6acf94a77c30ef96d6c6caac297908f7f6208"} +{"id": "train_42901", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28652e6ec20121ddc207e4c3c1010182a81e6f18"} +{"id": "train_42902", "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: 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", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1be850670f40d3413a0e247685bdd2467424de75"} +{"id": "train_42903", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fcb9771a9d0ec056b5f0b723d5818e4d391123a"} +{"id": "train_42904", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c355008bead2384bd3f8da2ddc612c2e532457a"} +{"id": "train_42905", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "security_gates", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "978c55fd9e1f24d1cbe6065777df1c604540e274"} +{"id": "train_42906", "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": ["tests_are_truth", "governance", "documentation", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8762852519ec5ff27bda31e9187493f6916650e"} +{"id": "train_42907", "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": ["repo_scale_reasoning", "auditability", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f148c5f1e660df010309f5cf065ed3a0de610877"} +{"id": "train_42908", "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: 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", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24dbc1edb129136045a290f6ccd480e97f03eb18"} +{"id": "train_42909", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fde16d633be6e66eb62077491b5e7450594e53f7"} +{"id": "train_42910", "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: 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": ["documentation", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f082c96a62e6f2b541496abaf25aa9e8eccabb01"} +{"id": "train_42911", "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: 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", "tooling", "auditability", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c19e7e7433b78a4999f6eaa90da284bc33e86ef0"} +{"id": "train_42912", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9b75da98f5499207c5072bdd1e3d1b3e547364d"} +{"id": "train_42913", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdef99cf093de1c8ab454cf07710d00c4a0f0b88"} +{"id": "train_42914", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b919603f2b9831f2df289efcf3b2e76e934630e6"} +{"id": "train_42915", "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: 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": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae5126d6b9cc3cef53c2a5a02e64f5605dcd13c1"} +{"id": "train_42916", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2c3827b087129ce7bd06834ccd89f9561852b82"} +{"id": "train_42917", "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": ["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", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6642da0a2409148c74c786d7e5b34a2391fb770d"} +{"id": "train_42918", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a1ad4babfc397e0bbfc0b6cdd96c1eb4966f39d"} +{"id": "train_42919", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "358482e25f7ecebc223a40e1f803f09bf9972f58"} +{"id": "train_42920", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["documentation", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e729f89816dce29fff86a4cc8980dd7438e072a8"} +{"id": "train_42921", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb25ce4cf392517bddba1ca6308147594267d21f"} +{"id": "train_42922", "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: 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", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f924935010f376e2ca28ec2cb6fdfee788e2b18"} +{"id": "train_42923", "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: 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", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6227aa65ad22429c8f7c106587b9f2c0aa88908"} +{"id": "train_42924", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0d7843dfee06a611b425770d0b8eb175e15f63f"} +{"id": "train_42925", "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: 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": ["tooling", "tests_are_truth", "governance", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c376fc618a087b21c52894d2e585883fea78ff96"} +{"id": "train_42926", "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: 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": ["evaluation_metrics", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9c2a5dd3f968108a0e1933dfc01642a455cdaa4"} +{"id": "train_42927", "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": ["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", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13c24bf7557a0d2b5ba2d48edd4249d135d424eb"} +{"id": "train_42928", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "635aeffc6f52fdb48fdb948b5ed9af400f572e17"} +{"id": "train_42929", "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: 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": ["ci_integration", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9832d3d9b20f3876b6262b512230a2150f46a3d"} +{"id": "train_42930", "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: 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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ac4021dbde693a457d587d845bda02834e9e34d"} +{"id": "train_42931", "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": ["evaluation_metrics", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "518c73c3e1f23c2a9df99e7dc2924cc2eedf25bf"} +{"id": "train_42932", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fc6480bffe5ed2bb8afb6043fdd82de1609151e"} +{"id": "train_42933", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "caf0321135f59059795dc86ce3f4f9be323fcc6b"} +{"id": "train_42934", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2acd4b754ccabde730e5d2ceca6f92e16783042f"} +{"id": "train_42935", "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: 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": ["evaluation_metrics", "ci_integration", "auditability", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ead41d3f05a95613a83bfc97dfa4900c89f0f76d"} +{"id": "train_42936", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09d2d0194f6c36d1b5604d4a35ff45f00cf19228"} +{"id": "train_42937", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5dc8a8b7f1707c86c4646b39f8a306ca2016577"} +{"id": "train_42938", "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": ["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", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "041d9f4ec29572e4e26c59ee12cc35d3cd022afd"} +{"id": "train_42939", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "tooling", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ade8268e6373475192fdea02f0957725049aa0b6"} +{"id": "train_42940", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3368801aa74963e193ae2f9648882c4eb3afc9ce"} +{"id": "train_42941", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["governance", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cf7bcf5ac1d9983f76b892b922601940de7a97a"} +{"id": "train_42942", "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: 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": ["security_gates", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a487c19341c1d0a1bf9f961e99137a30cd73ff88"} +{"id": "train_42943", "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": ["documentation", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "465f56f983c325a357db8c82121bdb2843382352"} +{"id": "train_42944", "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: 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", "governance", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e240484c94609a8bfba9abf411c01eb263c0271"} +{"id": "train_42945", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9ae39fc3210d9a99a9da7c3c712e4090fd88da6"} +{"id": "train_42946", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["reproducibility", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3a3f3b2ff97fa6f8d590d6bf8eb8ee0a6e17841"} +{"id": "train_42947", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57d2201d2b6424cca9ffe235805e4676a9bf5cc1"} +{"id": "train_42948", "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: 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", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fbc2b8ab0d76eb7651c0820a019ca58fcf461ce6"} +{"id": "train_42949", "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": ["security_gates", "auditability", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afee596d471c108ecb0c671d64f8c8ea1154daa5"} +{"id": "train_42950", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06ac2ac9e6ffab34cd827c2eb0180caee2c81b5e"} +{"id": "train_42951", "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: 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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "006c8046b4a14736666f198ce3830499e3c9a38b"} +{"id": "train_42952", "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": ["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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87640cb25da76df6692db20c671133b5ce60af1a"} +{"id": "train_42953", "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: 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", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86f2ae2427ac3ac5ce5ac47e7aec3bcb291807cd"} +{"id": "train_42954", "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": ["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", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f2f36e3dee24967b0766a84ed77d50b91164b88"} +{"id": "train_42955", "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": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f9a36b8c8f4ecb8247034921d79c94d936d30df"} +{"id": "train_42956", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5976639c522c4bcb35f7f7e15bf16b06ae114bdd"} +{"id": "train_42957", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53936f17f350a8310622695dddf65809ef3887c9"} +{"id": "train_42958", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de549bb6be831f8bce57d7e474424a27692eed69"} +{"id": "train_42959", "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: 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", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b77a11b151387660354745981aaa7bd4aaca794e"} +{"id": "train_42960", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8258a50b8f1dae2ce1c77c38aad0305145c2f9fc"} +{"id": "train_42961", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ed9f3598c4379cbefcf71c698e79591e2ea5b44"} +{"id": "train_42962", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d14d5b86e126931c7b1a64bf4ea5ecaaf384c0e"} +{"id": "train_42963", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "governance", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50db3145f7bed8175a3004d859cae6a9e74a9902"} +{"id": "train_42964", "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": ["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", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68085678c5dc3be9b15d5276a28963e3c79f67c2"} +{"id": "train_42965", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "cost_latency_tradeoffs", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60b8d5c8cc3b6b924399f1435a7c03cbe577b2c6"} +{"id": "train_42966", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fbac5cd5a13700f817f510ca8ece5ede7523cd9"} +{"id": "train_42967", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d39994ac566175c227631e77ee9252a9c4611224"} +{"id": "train_42968", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "reproducibility", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa82d614dea88d39983e08bfecf000d5b1386639"} +{"id": "train_42969", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "auditability", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35eef3c1299f959af9d8774e252a80b568394ee4"} +{"id": "train_42970", "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: 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", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3c12be4488b1eadd558d499efb790cdc50e219d"} +{"id": "train_42971", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00446218973c848d436207d8cb9bfb9c32f185eb"} +{"id": "train_42972", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fa508a9aaf69ff95b1f42e7e0b4e6d604d3d4c3"} +{"id": "train_42973", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e729f89816dce29fff86a4cc8980dd7438e072a8"} +{"id": "train_42974", "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: 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", "tooling", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49d41657b62777a0b248c4c598ab76d5ac4d83ba"} +{"id": "train_42975", "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": ["repo_scale_reasoning", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7f9adee30364074f2631100f36d757930959c37"} +{"id": "train_42976", "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": ["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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58f4419d58b3045e5593a2578a96a6e67f1e1078"} +{"id": "train_42977", "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": ["tooling", "cost_latency_tradeoffs", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "065d335d6e2646c5816962daf5e8b50475b62144"} +{"id": "train_42978", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7cc87ae79683495170804956536e7c5b2752ade5"} +{"id": "train_42979", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed8ac455a0465e9d356b46132675498ec5296d40"} +{"id": "train_42980", "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: 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", "reproducibility", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3927c7555c417d77e89f5e1353a1c8cabb90dab"} +{"id": "train_42981", "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: 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": ["reproducibility", "repo_scale_reasoning", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8964783491f9dbe404451de44bb7d60a6082f0e"} +{"id": "train_42982", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5e4d6dd7f61fbacee0391db6189fd0d7ac81877"} +{"id": "train_42983", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccb14d32989bf8e7e49d0e31219366f52e6e9730"} +{"id": "train_42984", "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": ["tests_are_truth", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f413e73d7821a0d3e2557a3e4d8b029e21a7224e"} +{"id": "train_42985", "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: 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": ["ci_integration", "auditability", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2221df414100b0556f01576d6ae1b9fee4be5238"} +{"id": "train_42986", "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: 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": ["cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1288a46a21a07eb536bb5cd66905623371b47275"} +{"id": "train_42987", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58685895c5a02dcd59b89fca86c3cff0379d428c"} +{"id": "train_42988", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75d5c5ff9d16fcd47bee6adcf58dea3ec3480b16"} +{"id": "train_42989", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "ci_integration", "tests_are_truth", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3e662a4977cb1451c61d4cc3041e464ae58b210"} +{"id": "train_42990", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "reproducibility", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e42461cbf49b45e7ae463ea3934ca0cbde3ecb1c"} +{"id": "train_42991", "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: 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", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecd0137f69c9c0bc25f4b08cdf161d0f1ba7fd65"} +{"id": "train_42992", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2efa457c08fb2a1c59514a1cc42724f3b62a345"} +{"id": "train_42993", "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": ["reproducibility", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0882456aa52f0aaf9004964a4dc439a3a4fac3d"} +{"id": "train_42994", "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": ["tooling", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39098818073d0ec670c9c9053b48ae13697766c3"} +{"id": "train_42995", "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": ["tooling", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b115a8c9161f348db1525e517dfd6a344f6e780"} +{"id": "train_42996", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9bea7a7e2206585be39b80008e8252597badf252"} +{"id": "train_42997", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb3132cb4b038d2cd542eaf3e1709fa07c7c9eff"} +{"id": "train_42998", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["tests_are_truth", "tooling", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78a240c619e759d8601f264b06d47066b46bcb3c"} +{"id": "train_42999", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b185884742440f41f0017609fef9962bfe8d62e"} +{"id": "train_43000", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07f821c8e4a7cbb74a79a34b02cc9ed67eacdbe7"} +{"id": "train_43001", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a92237c860070833392ab58aabfbdcdb10891578"} +{"id": "train_43002", "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": ["tooling", "documentation", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a40cb459224ae5640510babf1701ab8fda9c7841"} +{"id": "train_43003", "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": ["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", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ede2f21e135c847c3718a1120440138d7b9c9fd"} +{"id": "train_43004", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["repo_scale_reasoning", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c702830f27a0f4b564406364989b9fb9e201ae83"} +{"id": "train_43005", "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": ["governance", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "641df3573b0ccddfdb8f29c2285e28af48dfba5a"} +{"id": "train_43006", "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": ["security_gates", "cost_latency_tradeoffs", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af47f786a2f62667f44778f51b47dcba61559ae0"} +{"id": "train_43007", "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: 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": ["tooling", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ad02c4fc05ed72fbd52672717cce316db8fc96f"} +{"id": "train_43008", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["evaluation_metrics", "reproducibility", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5d0435432a99685a44bc8fce37180ba7f456a71"} +{"id": "train_43009", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "273810dda51ab2122b3bc7b12c8835c0066c22a0"} +{"id": "train_43010", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["security_gates", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0cf8dbba184c70066ddb781bbb700aee5d18610"} +{"id": "train_43011", "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: 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", "auditability", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d574fcf1251e8e2a71fed5111c23239c0ac60aaa"} +{"id": "train_43012", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14ca084878430fdf87c27d3e97c9cf37249dd173"} +{"id": "train_43013", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1786aa9b99a8a2e5c65405e47f50da8ec5a4310"} +{"id": "train_43014", "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: 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": ["tests_are_truth", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75adbc8e78274a5f19ff95fad2d764dd7834cc1d"} +{"id": "train_43015", "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": ["reproducibility", "auditability", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8deae43905bd3e9d2a1d4b216c5fef0b751b562f"} +{"id": "train_43016", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d69ec9441d573bb42e577274e6b7f66fac9f5dc5"} +{"id": "train_43017", "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": ["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", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9565440dbd6c245750596a137a40ebde11e6a617"} +{"id": "train_43018", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "257fc64b241e73c7b8b750f32b3793433b57d08a"} +{"id": "train_43019", "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: 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", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64c0e3f32ee5f33337ff452cb10d3fadc6d0ef6f"} +{"id": "train_43020", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c1bf2a1e6f0be7e7300cc962c9d19c31c07d358"} +{"id": "train_43021", "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: 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", "ci_integration", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71417bf1d95ba28ce3bb3106d3fbd119b7aa45c9"} +{"id": "train_43022", "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: 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", "reproducibility", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e11a9309c3ca2575499f87a7931989ed28f2946"} +{"id": "train_43023", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6bf647af8b637ec37b15a701e1c97a4d8c09a15"} +{"id": "train_43024", "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": ["auditability", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5245c44820611622330bfc2f59f24381ee16791b"} +{"id": "train_43025", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "002caf4ea726e6423c5bace628d7a93a89a69321"} +{"id": "train_43026", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ff3b4b79e5febfc0c6dcd37724c6f017350534b"} +{"id": "train_43027", "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": ["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", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b539e6f38883ab91d86f725984ab5e83aaf4dd7d"} +{"id": "train_43028", "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: 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": ["governance", "evaluation_metrics", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1908c7c21f0021f41e6294f919c6aefad7c10095"} +{"id": "train_43029", "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": ["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", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a48b88a0b02911058e605dc453945aa3885395de"} +{"id": "train_43030", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0717bbab8d779be708723acbad2e7ce1bdc87327"} +{"id": "train_43031", "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: 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": ["documentation", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "160f14673dd3353ede01c043efadc978c2f625fb"} +{"id": "train_43032", "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": ["evaluation_metrics", "documentation", "auditability", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d6304a6f536583ab57b342a59cf5a2f2feac519"} +{"id": "train_43033", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "149ea5e4261614034930ef1aa6c98e0ee41f25bd"} +{"id": "train_43034", "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: 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", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "abaf6989344bb05faa478e8deb962e21101f4ef3"} +{"id": "train_43035", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf745f55acbaa29f2753731e4d03e7a4d13a9b53"} +{"id": "train_43036", "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": ["security_gates", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c43f008023ac5502f461ab1c70821ecbf5eca79a"} +{"id": "train_43037", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bca70ad3e009fad7bac664144322f060d5561bd7"} +{"id": "train_43038", "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: 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": ["tests_are_truth", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "476eabee9df0d55d0e00f8fb049219293bf15f80"} +{"id": "train_43039", "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: 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": ["ci_integration", "evaluation_metrics", "documentation", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0979455f08781096c180d5cb055105c9922aecfb"} +{"id": "train_43040", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33420a14480fc4eb9620274c188c36dc1d4d536c"} +{"id": "train_43041", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f73596c29bf88e946d2cd87fd926462cf3d5c21"} +{"id": "train_43042", "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": ["cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed4f8ae03e72439d302ece9a9da0252c7ec58a25"} +{"id": "train_43043", "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: 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", "auditability", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e85b8008fb78fa3ba0bd6a5431bcc5b6027c08a"} +{"id": "train_43044", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb26d8741fe77855ad6427dd0ab8a004fb6bc195"} +{"id": "train_43045", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe98195c38de21b4bb6ac8d3992402c530c65b39"} +{"id": "train_43046", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "cost_latency_tradeoffs", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "511bb8241930aeb214d399bd67edceb5e2330d15"} +{"id": "train_43047", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["cost_latency_tradeoffs", "reproducibility", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "375627ae41815fc4fc3ac3e7cdee117698899079"} +{"id": "train_43048", "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": ["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", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cf546a5b0dab84c0c31f809fe1fcf48e28aa63d"} +{"id": "train_43049", "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": ["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", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5671f80633f13db5401223bd74652b73bbedcf61"} +{"id": "train_43050", "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: 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", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9c43397811f277df0c933b5b50c4742de3191fb"} +{"id": "train_43051", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88ec570038f1f7e84d56e0cdab7b3dd3ef6e8642"} +{"id": "train_43052", "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": ["repo_scale_reasoning", "auditability", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07b9c501e82548571c53a2ce8eafa22f6e042866"} +{"id": "train_43053", "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", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8a23a4050ee782d8a3539acfcf807d94fb7820e"} +{"id": "train_43054", "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: 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", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4b98d32a3a8e8ce589d56167cb247a9e846afe6"} +{"id": "train_43055", "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": ["security_gates", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6998a19a2ea7c54fb040966bd9c7b8b55ae0f0e5"} +{"id": "train_43056", "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: 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", "reproducibility", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e265113e0a53e4f981d00086becbb875e4b2a91e"} +{"id": "train_43057", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["tooling", "governance", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c3479788d4f730f354d28388301f1c8c064d3af"} +{"id": "train_43058", "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: 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", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b876897f400405725a89340583144868ab226752"} +{"id": "train_43059", "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", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fe14a714e2d923a836e9836a95e9713ca2caa4b"} +{"id": "train_43060", "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: 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", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9aa0a74fa914ba427822771ee0034af459d5dc7c"} +{"id": "train_43061", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "auditability", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dcc50792e3b7e70543ab287ca8999d1161b89d3"} +{"id": "train_43062", "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: 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": ["tooling", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8994642a323e119c8f57c8eb9f6e3c6e66b5a97a"} +{"id": "train_43063", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "676ba9f4b843ce6f5ac44c3eb8e00104e55ea5af"} +{"id": "train_43064", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["documentation", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b298aff8d924f5d39b0cbab0b416ceb56e933b37"} +{"id": "train_43065", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1b96ac5d600da89cafa50689473a73944bcb186"} +{"id": "train_43066", "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: 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": ["governance", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd660703052e00afb303b89aafdccc5550e57fe3"} +{"id": "train_43067", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5fad0c49208ccd198b9f40944244476bcfb6c44a"} +{"id": "train_43068", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "security_gates", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f555c4e34807e78316cc6eac2ec78d88d3c9ecc5"} +{"id": "train_43069", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "966c0fec35b08f38b3e77521bf27f25c99563719"} +{"id": "train_43070", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["governance", "evaluation_metrics", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc75b5c142e122b6056eda39f7a1675eb27c9bde"} +{"id": "train_43071", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tests_are_truth", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5f986eb4dcd9240b574df50af57f7c829e9f5fd"} +{"id": "train_43072", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88ec570038f1f7e84d56e0cdab7b3dd3ef6e8642"} +{"id": "train_43073", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98cc578131a7971f85edc852e3d6d0d3c9937da4"} +{"id": "train_43074", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c21ac80968d04d45684403df130cb160b32d2179"} +{"id": "train_43075", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40e96bcbb2497bf8f24899fce3d2026deab70817"} +{"id": "train_43076", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "tests_are_truth", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e9a1e3f27f977d2c8c69b4de815bc30a84c3b46"} +{"id": "train_43077", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a502b6ab819d22d969b5f581a2e449ca60a4402"} +{"id": "train_43078", "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: 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", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4bbc41f0da1d7d7a4d6867babeee4a8db05b57ef"} +{"id": "train_43079", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0962e4cd11a2af01aa14162ce958bf3a779ae81f"} +{"id": "train_43080", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3502d4c9de766d393ff4a309207ba372a19de305"} +{"id": "train_43081", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e3b935c76640d6457feec55f9b2e01669870379"} +{"id": "train_43082", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58c35bc22bbec02faaa6edfed7282be7a161baac"} +{"id": "train_43083", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "288bb1b86c4cb1b7c6e0b8c39103b2309fbd6385"} +{"id": "train_43084", "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": ["auditability", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a18377aa252fd9b10874009b0dfb8dad207ece41"} +{"id": "train_43085", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4b3257b50cfcdf01d31797d0748262ca5a22913"} +{"id": "train_43086", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "651ff55b58cdea360fe77409ea0a3d04ab450eab"} +{"id": "train_43087", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f75193d1bd21d4419b6995d38305967e60240c5"} +{"id": "train_43088", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9e355a5a9600c4b053f02fdb0070426237b5652"} +{"id": "train_43089", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36a41c92140584c703a57cbbd55be3b6d7db3e24"} +{"id": "train_43090", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4086f4c5614e350325dd53a62802a26c3a5a75e"} +{"id": "train_43091", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tests_are_truth", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55949926a17e901f712b595e8cca73b91722f6e4"} +{"id": "train_43092", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "385a393b69d55cd1d57e32576a1635abce39d6b4"} +{"id": "train_43093", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready 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", "auditability", "governance", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1be08020ea60e1bf21bc98044e89e1ad2d8f7939"} +{"id": "train_43094", "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: 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": ["governance", "auditability", "documentation", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77d686e2b208c7f3289d2b734fc5bd81f1fe0d44"} +{"id": "train_43095", "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: 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", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daf7eab2e33aefc9b451752d59a94f267bd19347"} +{"id": "train_43096", "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": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16d7cc0aab968b3845ac692a7c25177462e86afd"} +{"id": "train_43097", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1aa8fe4c1f39f8a7a0827e65b5a3a501d975e896"} +{"id": "train_43098", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf283fe0a5fb3fd359f956adcc79cb04072d23f6"} +{"id": "train_43099", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "688786229379a1b00bccc773e9de581a96846db7"} +{"id": "train_43100", "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: 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": ["documentation", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c0084ecd9d31bc05dc73bf6dfbef710b9a330f28"} +{"id": "train_43101", "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": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c476d147ead669d64679c8cd2aa66f554291abb6"} +{"id": "train_43102", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37cebf5d2989ca1c361f29b0349a5bc52c73f0cb"} +{"id": "train_43103", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55ded0c70438c26998c4e50c44eb1c2970c0590f"} +{"id": "train_43104", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebdac0c8a15fdbe143d21f58aad2d5c0bf1266b9"} +{"id": "train_43105", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55675f2f4816013fd97836ba4d6a79953d8753d8"} +{"id": "train_43106", "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: 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": ["security_gates", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5580760eb597db7143de488d91f1c2463e4950d4"} +{"id": "train_43107", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ac6735f9fd47b52756b3761c9c722a0e208628c"} +{"id": "train_43108", "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": ["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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae8a40f6e80f927f71850df45e1c105541a3a2cf"} +{"id": "train_43109", "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: 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", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64fa7445deafa91fa40d65c9d48fd2af8f54268c"} +{"id": "train_43110", "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": ["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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42bf0b6efa0a1261e47fc2d59f52245069189811"} +{"id": "train_43111", "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": ["auditability", "reproducibility", "tooling", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab912c4215bdab613bd01cc34f61601546ce56aa"} +{"id": "train_43112", "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", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1546dc3dddf616c996131e5d44e206c9bfaf6c5"} +{"id": "train_43113", "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", "documentation", "tooling", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a04dbc0bb74e8e9f54a0206361c00b1d46a4f65"} +{"id": "train_43114", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d6f73d29e139533d7d6bcf23e6884c0431e008a"} +{"id": "train_43115", "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: 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", "security_gates", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e7a04f2ab28b84a7bf0f437dabe73e68bfb72d2"} +{"id": "train_43116", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11523e259300fe87aafb75fad9a80bb2a744fc83"} +{"id": "train_43117", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["reproducibility", "tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90415e0ce59005266deeb6826cc2683ff75ee16d"} +{"id": "train_43118", "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: 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", "security_gates", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb3a4564a7be6f0a89870c7ec35d8df94d63a2d7"} +{"id": "train_43119", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "auditability", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "676f00d2adb06f08b253ef305c8208424d510fb0"} +{"id": "train_43120", "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": ["auditability", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a490765279bb9f6a4f9d63e5d28b021052cad560"} +{"id": "train_43121", "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: 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": ["tests_are_truth", "governance", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c92bd6428e3ac80369bb5fad4bbc2a6b977bcd3"} +{"id": "train_43122", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a112c35499765c582d2e591467c83dce5398c648"} +{"id": "train_43123", "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: 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": ["ci_integration", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b51b4ee0e802c8069c775533433d103e39ca260"} +{"id": "train_43124", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["evaluation_metrics", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f6b2da27cde63a8d5076211c3b671526b2daa32"} +{"id": "train_43125", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4be97a35f487479b7880fda326a10d0e139d004"} +{"id": "train_43126", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "security_gates", "ci_integration", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ad4286e6a79a849275579fae3fd5741a49d89f9"} +{"id": "train_43127", "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": ["cost_latency_tradeoffs", "governance", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "863363bb0116dc036835424f8a9f51036b513fdb"} +{"id": "train_43128", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b5e3b7453fd1a6f0b5f18e639cff148e38e2cb3"} +{"id": "train_43129", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33b89348338c3fd76334ecb77f5f7680b3a40239"} +{"id": "train_43130", "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: 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", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "007a13c606ccccf283138adb87385e2852c9abf4"} +{"id": "train_43131", "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: 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", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d53b582b3acb1ea6837e47afca2fb595ba58e52"} +{"id": "train_43132", "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": ["security_gates", "evaluation_metrics", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "568f52ccb5e4f4a08e3223f3e56c2fad0de22520"} +{"id": "train_43133", "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": ["documentation", "security_gates", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8112abec3e76860411f093ddabed2c4f5bf74bf"} +{"id": "train_43134", "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: 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", "evaluation_metrics", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "079cefc92567c85c26d1d61b7903b0abd9d5d5da"} +{"id": "train_43135", "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: 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", "auditability", "governance", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77dad616e7dae73cc8db86293871bea9056f1983"} +{"id": "train_43136", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f396a87f7b186f37e17e0704c8ca43c330be1a2f"} +{"id": "train_43137", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4438eb89526d1cca8b2fd4313143d431750bb15f"} +{"id": "train_43138", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["documentation", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4abd6d6a3ded36956166150ae96e74834474d502"} +{"id": "train_43139", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae85fae3b789053c4c5bca668f2f8f714a865be4"} +{"id": "train_43140", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "495417012fd3f2d6640108af8d9e386d9f8e73ea"} +{"id": "train_43141", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b57c89c3990811fbce53348e889e7c736556deb5"} +{"id": "train_43142", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f06c138e22dbd4cf20410a71dc024166cfad01a2"} +{"id": "train_43143", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86d13ef34d55cf2d9e1c0420ca7718dbfa1b0b58"} +{"id": "train_43144", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4bbc83a77aaa5f81fc81648eab21b5cbf5e5b18"} +{"id": "train_43145", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd06a59dc9445b512857c9e9922c05f57f984205"} +{"id": "train_43146", "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": ["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", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f82276d99ae592fb16addd9df7c88f70ef1db2ae"} +{"id": "train_43147", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["reproducibility", "security_gates", "ci_integration", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ade4f69f450e73c4d9a0f6b660e284ea5b113df"} +{"id": "train_43148", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "security_gates", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1acf6b40ecf40315cba9f1402118ad4f26388b76"} +{"id": "train_43149", "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: 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": ["evaluation_metrics", "security_gates", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96eaba6c27ecd6469706a7b024a803a7ceff369e"} +{"id": "train_43150", "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": ["tooling", "reproducibility", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55a869b6853401241f18c95dc0fe0703306da9d5"} +{"id": "train_43151", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dadec011c37f06b15fcf3c17a5e2410f3414f05"} +{"id": "train_43152", "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: 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", "reproducibility", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52a9eb49f1876e19f373d8982152e268de47dde0"} +{"id": "train_43153", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1287cee38742f0356328851077c2f99a0a8c10b2"} +{"id": "train_43154", "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": ["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", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4538c35740dda6c0e9116acc48767280d21b489e"} +{"id": "train_43155", "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: 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": ["auditability", "reproducibility", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d60373d57b0de5d0020de27aca020e388f1b1f7a"} +{"id": "train_43156", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "faf77ae3776a978ed9ae55c705199682dfd7f572"} +{"id": "train_43157", "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": ["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", "reproducibility", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b639c5452a90e7b11e47b50989f186eb7db8f68a"} +{"id": "train_43158", "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: 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", "security_gates", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed102da6b980a704e63a78db8bdc7ef66e3addce"} +{"id": "train_43159", "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: 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": ["security_gates", "documentation", "reproducibility", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "882f83b546a7264dafd9b26949c5ff6070f072ec"} +{"id": "train_43160", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f22120f22d86ac787942beff4f269a138136bc18"} +{"id": "train_43161", "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": ["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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3f614984ec3bf8b56e1b93c53a5bb433eecbd9c"} +{"id": "train_43162", "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": ["tests_are_truth", "repo_scale_reasoning", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "193f9df08bac464b03d79b4bb2e6d66dfcec981f"} +{"id": "train_43163", "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: 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": ["reproducibility", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3672d6eaee351254c3935b9cbd87db24b21ec1b"} +{"id": "train_43164", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "980ee78a5ce722d95faf17d2b42044768c2ecc4e"} +{"id": "train_43165", "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: 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", "tooling", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f368addcb0999a645dccfd18d00cc429f96b2a41"} +{"id": "train_43166", "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: 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", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad47d8d7cbb6a574c5a574fd52d007bfb2b63ebe"} +{"id": "train_43167", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afc05ad07afe245a035947125a03667895386408"} +{"id": "train_43168", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bc5c9159e76b5fbe7f7911902f5311d1b8791b9"} +{"id": "train_43169", "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: 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", "tooling", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a1944ce4e90b6102cb5298900ecd670022c9a34"} +{"id": "train_43170", "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": ["tooling", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e57d2f2b459c32d7f2994a3c12021c7421490a2"} +{"id": "train_43171", "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: 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", "governance", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5802a0f98c087a0c20648ae2ed809b286c4ee5f1"} +{"id": "train_43172", "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": ["evaluation_metrics", "cost_latency_tradeoffs", "documentation", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58ead36d428e5b5364fc461f70c9c4823e2b17f1"} +{"id": "train_43173", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "499b05778cf0bf4bd3dc0cb7c6c590e20cb2e974"} +{"id": "train_43174", "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: 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": ["evaluation_metrics", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2846de63750d7dffc9f3ae1e989c00419b3369f4"} +{"id": "train_43175", "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: 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", "repo_scale_reasoning", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0126cac63cc7a490747de5c25c305e1b57a47ebc"} +{"id": "train_43176", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ad33d3d232cd737e1e6ec7f6f79b17eb225563a"} +{"id": "train_43177", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6099082e5eba6c30801bc00be459ba3e9e07fb8a"} +{"id": "train_43178", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7381b00502c7d4e519a5cafef2ac1ca70fa9e41e"} +{"id": "train_43179", "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: 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": ["tooling", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "996341ae10d5eb571a7b256df1b044ec3eb015cf"} +{"id": "train_43180", "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": ["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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cca97d51283b3989a800d4f9a16fcd96ad6a1e88"} +{"id": "train_43181", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "595dad8f70743a09e8fcbcce55332dd2b49aa0bc"} +{"id": "train_43182", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8feb00e71854b4e5163b6d07ed523d908c92a4b9"} +{"id": "train_43183", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bf3dd125247e36e7d06813cdd0004153083779e"} +{"id": "train_43184", "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": ["tooling", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc47c06cf24af6d938547b59f8b2645aee812b1d"} +{"id": "train_43185", "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": ["auditability", "reproducibility", "ci_integration", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc21f38243be8b6430c8b7159a8483989294ee09"} +{"id": "train_43186", "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", "reproducibility", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6346e8bcc641d707539224baf1b8f043d47d21c3"} +{"id": "train_43187", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97a3a899314ae5561eb57a5345fe5253488f20a3"} +{"id": "train_43188", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee8fdee36af95e87ecf48c95013eda603a5e8cdf"} +{"id": "train_43189", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "security_gates", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d274550ce93444b03fb6a88f2aa27048529bf6c3"} +{"id": "train_43190", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cfa467cf7a919dfaad892e589d5da6ace8daa46"} +{"id": "train_43191", "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: 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": ["evaluation_metrics", "tests_are_truth", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "015a5ae8e239b7fd9b69c6bf5cef8df703795ec7"} +{"id": "train_43192", "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: 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": ["security_gates", "evaluation_metrics", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b23deb30c46616b98ad199c00744c5096008ce39"} +{"id": "train_43193", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b07a601d688084d6d90bd688fbc689c6f655aea6"} +{"id": "train_43194", "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": ["reproducibility", "tests_are_truth", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b360361ef83ab5ddefdbefa10b8e83892d0f42a5"} +{"id": "train_43195", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40c9a1c539a8c33009e994aeeb15761943212b6c"} +{"id": "train_43196", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "293fd1f09857e73612b114fc7f76789899057c11"} +{"id": "train_43197", "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: 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": ["tests_are_truth", "tooling", "auditability", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d8c81c7dbc84a94f3a82c054c4db1cf100e8aa5"} +{"id": "train_43198", "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": ["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", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab9c3c5027ac1e027dc8ec3cc2bfd3022a0ac6a7"} +{"id": "train_43199", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c56237399e36a54cab96f63d4249e0d24414929"} +{"id": "train_43200", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "288bb1b86c4cb1b7c6e0b8c39103b2309fbd6385"} +{"id": "train_43201", "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": ["tests_are_truth", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae8411a795d69207c0953bf58cbad41f18c5156c"} +{"id": "train_43202", "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": ["auditability", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34b5831028228e32d01c8411779ad9b708d64b3e"} +{"id": "train_43203", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2dccd481e4f255b9ae1ffdfde0554af084fbce1d"} +{"id": "train_43204", "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": ["tests_are_truth", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b96a8f42700318886c27ef50bd97877328a11002"} +{"id": "train_43205", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e3e366c9f872de913c06258071f11143b5b4197"} +{"id": "train_43206", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6f11c7a7432e7425d0865efaf93bdaab1fb9766"} +{"id": "train_43207", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a60db7b156b49efe1c5378fc75cdb4fda3fb9d6"} +{"id": "train_43208", "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: 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": ["reproducibility", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf92ba018b2bcd34968942dfc21cae3e63448adf"} +{"id": "train_43209", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d559bf3f91286992d55d6679105f32afcdbd13b4"} +{"id": "train_43210", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e63c563aa75db0d4e93207ec32c87819810dac1e"} +{"id": "train_43211", "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", "auditability", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f0c628a7240fb99ae81c8740fe4eb1910e93c14"} +{"id": "train_43212", "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: 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": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7b3ccf77d2dd06bc571b535d83a16b142f40b27"} +{"id": "train_43213", "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: 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": ["tooling", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f75b9ef52f386c951da4d2369c5c4f4c2540f21"} +{"id": "train_43214", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b43c5aafc058bcab5e0ad582ac91590dc3fb5979"} +{"id": "train_43215", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e8df88610e8112604d200feae7a67f9c9ff1f32"} +{"id": "train_43216", "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: 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", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1115301701af81904a8bc8322df46bd8e6d2e46d"} +{"id": "train_43217", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85379bf275937e1453418dd428cf445e243c9f73"} +{"id": "train_43218", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d4c3dda36faad73f4b891723ec7182d869ddbb6"} +{"id": "train_43219", "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: 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", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f2fa10271f46b3ec82eedc94be90da5bd2dd07b"} +{"id": "train_43220", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdee0bfc761cbdc08513d0ece086225f1a50c734"} +{"id": "train_43221", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "894ae7c212462cb0ce4d74137b19c2bc64408f28"} +{"id": "train_43222", "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: 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", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4317206211ec00d236e7a5b81c12393a59206c89"} +{"id": "train_43223", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71e29496cdba2867c2fe9ca0481ce2b9b06306b1"} +{"id": "train_43224", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9db365cad2e0283b4e72d8ab9222e1fe9e37500d"} +{"id": "train_43225", "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: 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": ["documentation", "ci_integration", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "647025767d068a73bc55e13c2afd939d5e449972"} +{"id": "train_43226", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8512390d43701d3230eef220d09d031100fc4f46"} +{"id": "train_43227", "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": ["repo_scale_reasoning", "governance", "security_gates", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42f212a06aaaab64c2ba0ba61c715798d0910629"} +{"id": "train_43228", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27d1b37b67ab1fc8745da5ac5f9c823ec28e002f"} +{"id": "train_43229", "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": ["reproducibility", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "314ed10ff5485d1b4a7945291aa669e4608bc3f4"} +{"id": "train_43230", "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: 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", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56852ad7466b7bcfb5093063cfe3deb9cd8e25b6"} +{"id": "train_43231", "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: 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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "005d15c1e24b23f56682f4fbd0c97ea9e95eb055"} +{"id": "train_43232", "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: 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": ["documentation", "auditability", "ci_integration", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d29d8dd5058eedb6b00322ffb13e0df6fac9545"} +{"id": "train_43233", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cd3ee1bd9a066c4e4b79029aa777443a422be0a"} +{"id": "train_43234", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be91e4ac5cf08a2cce1f29c1ca110b0821f91ee5"} +{"id": "train_43235", "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: 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", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "233ef8089b9121ab335083d5a7f0ed663b6e75ed"} +{"id": "train_43236", "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": ["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", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7f431d7ad246451e014e779fcaa963220d7c032"} +{"id": "train_43237", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["documentation", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ff89cca5ebe3fd80cf669d899dc5ffae2870e8"} +{"id": "train_43238", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1e8e18f14bbcdc00506991293dd953601162d16"} +{"id": "train_43239", "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": ["reproducibility", "ci_integration", "tooling", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a2099dfd1a4e6a29d066c33962c7693d9b07fb1"} +{"id": "train_43240", "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: 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", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18954c6d2a1dd82d6298c761f94ef6c0bef06935"} +{"id": "train_43241", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12fe1d0a85373b8fc52d2aa1ac44736924633daa"} +{"id": "train_43242", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62d15cfe603cec62790d992f15f83ad853bcaff1"} +{"id": "train_43243", "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: 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", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9fd53d56fc098f0dbdad0ef25e9bec0ee49e3fa"} +{"id": "train_43244", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cff003dbab5f1aeec44c12c761c270d4fe9e694c"} +{"id": "train_43245", "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: 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", "governance", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "711bee2df4646b1ec1b57bc665d30a214444cb55"} +{"id": "train_43246", "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": ["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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f2be444f3bfcf11185897beaad796213672a984"} +{"id": "train_43247", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68889f452f3515263041ae7dffc814150f88562a"} +{"id": "train_43248", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab3d20cf307e104a7684b55fab3e9c0d6f8c8475"} +{"id": "train_43249", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a95b967bd2368d72f2ed2480d55afc2bc2b27a30"} +{"id": "train_43250", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75a7ed94cc1e1785cfe53d261aefe078bbfaa1d5"} +{"id": "train_43251", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bacfd6f4fbef560a86af2fb24a52646cede40d2"} +{"id": "train_43252", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35c8909978ba58f08ba6c93945af022e633fda4a"} +{"id": "train_43253", "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: 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": ["repo_scale_reasoning", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4347564c0c7e245c1f8eda820d7baa1173aaa170"} +{"id": "train_43254", "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: 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", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d501077c4759728d5fbe9ddfcaece960fa96faa"} +{"id": "train_43255", "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: 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", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08d517b3b53f278623901f963f3c5ba7a65e17e8"} +{"id": "train_43256", "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": ["security_gates", "repo_scale_reasoning", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96d80731f3f93e943fb572501ad7284b618c73aa"} +{"id": "train_43257", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bc4869fab869186d1a5e4d07b5f6f9ed6435cee"} +{"id": "train_43258", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f07f7fdfa5e7876ac1d1dc9b88314379120ba1a"} +{"id": "train_43259", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "810cfbd4582ed7a4ae03582a3de8ed147a4e0d2d"} +{"id": "train_43260", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "security_gates", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3147673bc89f032149880601e00246c691a4847"} +{"id": "train_43261", "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: 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": ["repo_scale_reasoning", "reproducibility", "auditability", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf3884f5bf7dbffc6e0d71fbfd359cb8ddd18471"} +{"id": "train_43262", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10e5743ae72b83a433b4efe991538212db6f2294"} +{"id": "train_43263", "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": ["tests_are_truth", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1c1d1a564401eee99768665e8b3a55ae3250d56"} +{"id": "train_43264", "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: 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": ["documentation", "evaluation_metrics", "security_gates", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3225d57a21649fab1b05e6e7cf5ec82ebee88a3"} +{"id": "train_43265", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "834763ac97c15740ca5111141a2a28fb20cf77fc"} +{"id": "train_43266", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "551e3bd57fa5d809adecd38f1505de7877bfd486"} +{"id": "train_43267", "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", "tooling", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c63cba48739fbac651ba47f59dad4536680b3f55"} +{"id": "train_43268", "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": ["reproducibility", "tooling", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48e91fd54557a76e782766945ce6400965e5f833"} +{"id": "train_43269", "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: 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", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "573dceb0a72f501090321fc5a328d0f114ede93c"} +{"id": "train_43270", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "ci_integration", "documentation"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "337f915f3daa47a25ec36620d50af97e5c19a4bb"} +{"id": "train_43271", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a86aa65b4dcf0ca85131eac93723b2738fedbb5"} +{"id": "train_43272", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "auditability", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3892395ce89d09a1e7c5010b4cbc6686102b3ec5"} +{"id": "train_43273", "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: 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": ["documentation", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2aced7bf3519af046e9fd4d810c352c733d86730"} +{"id": "train_43274", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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": ["governance", "auditability", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dedbc05b8953a335b77af727114245f8f4af63c6"} +{"id": "train_43275", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "security_gates", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4d9d29c043432339e6bd36f5db59f1deaec8d63"} +{"id": "train_43276", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["repo_scale_reasoning", "governance", "auditability", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5d0435432a99685a44bc8fce37180ba7f456a71"} +{"id": "train_43277", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["auditability", "tooling", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8681fc1d7512a29887208e4c6c6930a65327afe0"} +{"id": "train_43278", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca62baf89cdef451610b94a544971678635c14f3"} +{"id": "train_43279", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7173808b51a82c6c25d581102c49183f6755fee"} +{"id": "train_43280", "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": ["ci_integration", "security_gates", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b98c4e644f1de83a2c1c8104fae19003cb0a94c2"} +{"id": "train_43281", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d5bff36966c74b0c520f3f79054e4eef23e0766"} +{"id": "train_43282", "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: 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": ["reproducibility", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c09f9b7537609b31beecc3f81ff8ae165e399b4"} +{"id": "train_43283", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ceaea7dca75451790b9d650109ed26ddce17c153"} +{"id": "train_43284", "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: 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": ["repo_scale_reasoning", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87ab78c04ee8f020d805e2b8f6e8aba94c4fc490"} +{"id": "train_43285", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74509bd6e2c58b31208a0a085fdc7ab273b647f6"} +{"id": "train_43286", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc4d75ef0267b893b9a67b60ffc642475e9de2eb"} +{"id": "train_43287", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83a67c6720eb103c2637020132a7990385082d5c"} +{"id": "train_43288", "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: 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": ["evaluation_metrics", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b60fc2e6ad024f9d15473a8cc1e4e96b60a167ec"} +{"id": "train_43289", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f4f56b8a8690cc056e34f9649527f340ee284f1"} +{"id": "train_43290", "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: 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": ["repo_scale_reasoning", "tooling", "documentation", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca25e2a26cfb8146edfcaab0ec0cfe06a7e8642e"} +{"id": "train_43291", "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: 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", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57d5bea64ae3ddd318ac3d296fd45768c782715b"} +{"id": "train_43292", "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": ["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", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1137d3adce98b56bee53c9205d61849703415b1e"} +{"id": "train_43293", "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": ["repo_scale_reasoning", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d7eb4eea714c2a866f28099e658b6ac04c854da"} +{"id": "train_43294", "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: 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", "tooling", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9965e054b29f4825da4f8aaf6bce22d5b9fa6e2e"} +{"id": "train_43295", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24bdd59ebfc7fef3b8032ffd20dac3ad8a101db3"} +{"id": "train_43296", "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: 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", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "936ba14a8c15b8a38d701ed458969fdd7f7a3b34"} +{"id": "train_43297", "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: 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": ["ci_integration", "repo_scale_reasoning", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b975bca20c862800a640c80a39b4c2c38729e73"} +{"id": "train_43298", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6259155f769d9ac7330b5fffc72c1b254ec9bbe5"} +{"id": "train_43299", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["governance", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b60d45107dd25bc8c01415d8a4f508fd6321e795"} +{"id": "train_43300", "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": ["ci_integration", "documentation", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb7d66befaa9163cd0e6df596a9375f049953c3b"} +{"id": "train_43301", "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: 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": ["ci_integration", "governance", "tooling", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41fae4108a70e3237d16fb0cb5ec82b2ac17d7dd"} +{"id": "train_43302", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c19c331839a1d43bfea287440a822de1d55ae65"} +{"id": "train_43303", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df6641ba993530e1a2974846c5728b72606dc019"} +{"id": "train_43304", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["reproducibility", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c279efbd9dbc40c3860d63e43619679ac35979a7"} +{"id": "train_43305", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d377b7895343ec673ea1a2ed040c7c410429652d"} +{"id": "train_43306", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["repo_scale_reasoning", "auditability", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1155089bf70a72a09df0e3acb9748e71f3a3941"} +{"id": "train_43307", "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": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8c1e72a2a4a2aed55324ef6e385a210b6e7e564"} +{"id": "train_43308", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "042e7a645ab11133703c8d67a7b73f3ba38835e0"} +{"id": "train_43309", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "reproducibility", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7794fc749f10f87c123ca98c9ae1eb48ab0ff776"} +{"id": "train_43310", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5009519b3e1c332c2fe5e4804c9f1e7a46ba9a63"} +{"id": "train_43311", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9ee809ca90fa5ee163872bbefa52d0afba37536"} +{"id": "train_43312", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c73d2d3cadedca198355931921885870b6bd508c"} +{"id": "train_43313", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fe5a0d05aa7ee58cbadc84e45545526977c2187"} +{"id": "train_43314", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcb4bed5fde2e2c0c5937e43c4fa2c7cac64365a"} +{"id": "train_43315", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "935958c3ec9bef31e7bf49b877f400d2e34497ee"} +{"id": "train_43316", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "992655cce64f5bb11da52fd387d1178f25ce1f9a"} +{"id": "train_43317", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "auditability"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcea988881d8a8b8d846e549fa4ff72b51cbec70"} +{"id": "train_43318", "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": ["reproducibility", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a914947c923543b8099bc5a56769a5170ca7ee01"} +{"id": "train_43319", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "171306bc10e2d9deaa462f277f42b07733c473de"} +{"id": "train_43320", "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": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59d974c5117a9486a7345296f7759fc19a6c32ed"} +{"id": "train_43321", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f2b72c90e779c385fa9f313a6d8a630a9f75ec6"} +{"id": "train_43322", "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": ["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", "documentation", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43d235f0578dcc347e1a9a73ae6d3e89d9e1cf2d"} +{"id": "train_43323", "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: 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", "documentation", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63fed22b3a36090ec8053f15edff8d20540e19de"} +{"id": "train_43324", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2195a3ef034e70ab88ac480976548fe07b0e77d9"} +{"id": "train_43325", "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: 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", "cost_latency_tradeoffs", "auditability", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29012cc1229e3c87136696667e65f648fb476923"} +{"id": "train_43326", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["repo_scale_reasoning", "governance", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6d5556eb89ab379fa0609b3f06f181c73b71edb"} +{"id": "train_43327", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd73b69b29c77625c57636cd09485b007894a4e4"} +{"id": "train_43328", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1955b7e1665d0c3604d54a9d8cf930da03a20ab2"} +{"id": "train_43329", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a73388f59d83b37e1409b234ace6729783d3ff3c"} +{"id": "train_43330", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f5aa7f6f9908f0684077bca062e0e8a86f14286"} +{"id": "train_43331", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c99db5669b1592b572f35ee83591b1d31e1313e"} +{"id": "train_43332", "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: 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": ["reproducibility", "auditability", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf9fa46914c2bc2c4be469370a449253497be01c"} +{"id": "train_43333", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tests_are_truth", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8146f329db9697bd54f5a0b9406427bc3622b177"} +{"id": "train_43334", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "security_gates", "tooling"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccef1d1f626a6fe26eb5bc298aba1fec40084957"} +{"id": "train_43335", "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: 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": ["evaluation_metrics", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d521b7d562aed5dc80dd21ecc03fae0b6a88572"} +{"id": "train_43336", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7581fb642a66aea84b3e47ff12f065277a1fbf29"} +{"id": "train_43337", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "815f8814e815ff8ee693c5558af2a2d05ddb5144"} +{"id": "train_43338", "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": ["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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f2489f3784291151dfb2738b7b5f409a51a05b7"} +{"id": "train_43339", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a5d8ca1b5bb22b91b340d835a322b7e252a46e5"} +{"id": "train_43340", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27573c7984f89ac0dc3128319022fa741486ed8c"} +{"id": "train_43341", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25b3752fe34f428df7c1010316a48930947a6157"} +{"id": "train_43342", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f61641a8fa993ec16835a13298383164c8c5cc62"} +{"id": "train_43343", "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": ["auditability", "governance", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d222fd7b4585ea90108a442f925b8a60d660ee0"} +{"id": "train_43344", "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: 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", "tooling", "documentation", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a6b90254db202dfc67336b385eff18b1e288d3e"} +{"id": "train_43345", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25d41920df1c4806bf8e5b34d21157dbda013b20"} +{"id": "train_43346", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95bdcdeaff5ea53803f16b12b1a6c1b11d122292"} +{"id": "train_43347", "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": ["tooling", "ci_integration", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86f4d06124b09c086c1abceddf458e544f2daed0"} +{"id": "train_43348", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01c5277e3cdc604c09cd1b85b5fcbe45fc649d54"} +{"id": "train_43349", "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: 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", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b015e90359f78a9895cd7d42e3ea3e1c46460bd"} +{"id": "train_43350", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75217e3a9ce42e8d70ce6a6bd99d20fc9ed23cd6"} +{"id": "train_43351", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f09a9b566cbedb7ad21de1269b5dc2e11d3b61f"} +{"id": "train_43352", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "documentation", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e34cca04c25e964cbc0c3701ee8d481f43c66dc"} +{"id": "train_43353", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91259d3d84f0473a6cb67b891e4514d0fe309985"} +{"id": "train_43354", "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: 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", "documentation", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33f60a2c4aaaf6e0edcf53548bd52bf7c084c56d"} +{"id": "train_43355", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["evaluation_metrics", "auditability", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4348e2ace761aaa5df8acf931d24fd2639d928b"} +{"id": "train_43356", "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: 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": ["repo_scale_reasoning", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15593ef14303f584fd59209bb48a68260bf4abd0"} +{"id": "train_43357", "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": ["ci_integration", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf1bde98889720e6ffca5a34fbbaa8bddcce58f6"} +{"id": "train_43358", "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: 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": ["tooling", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c9c0be4af001d199df74f825aa290f022bdf280"} +{"id": "train_43359", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a50b4d8d1dcf4e3cfc90b0663ae665ebbc6823fa"} +{"id": "train_43360", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "governance", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b777104820a39ebdbc2cd9a44cac053d94166f5"} +{"id": "train_43361", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["security_gates", "governance", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eaadc7013d6bace09e99e2af274da91ebcb48405"} +{"id": "train_43362", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1629f3e197914ea01aa77baa6e1c149dff9a339"} +{"id": "train_43363", "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: 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", "tooling", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80c93df682ce284933ad9eff2b6cd4d94ee9e0a1"} +{"id": "train_43364", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b04d5226090ea7cde1d60550632afa048db924d2"} +{"id": "train_43365", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f19097df2225df427fd55518469cf4f01f0c686"} +{"id": "train_43366", "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": ["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", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c64611c81f77f9277ef7a93fafbff2fd86c23ce"} +{"id": "train_43367", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "tests_are_truth", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b30019b198a402048bb7d949a3ded373acf4844e"} +{"id": "train_43368", "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: 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", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cebff8542de10913e60e85b148459a72430de49"} +{"id": "train_43369", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d411ddc04151212bb65c39125db5e16bb7b2401d"} +{"id": "train_43370", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["security_gates", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "612da9b11053ab61e7ef44d26435d87f88e031ad"} +{"id": "train_43371", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8d2c3aa8ff0253f1ffa119b38dc8913694344fa"} +{"id": "train_43372", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "security_gates", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cacc0246dfd6a7bd96713024e3f9161756844728"} +{"id": "train_43373", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "governance", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "957048ba966c37be888d2251ad2e47c0bc804257"} +{"id": "train_43374", "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: 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": ["security_gates", "documentation", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "935c48beccd1ff28e3a31323a03f352898431f60"} +{"id": "train_43375", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a758a82c6dc8cb9b82751d37ac927a338768466"} +{"id": "train_43376", "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: 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": ["auditability", "documentation", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d9fb73fa331f34a8f5d55720b8dc708e1e0f1f8"} +{"id": "train_43377", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "JavaScript", "developer_needs": ["governance", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab2901dbe12a8494cc0e51ae4650f2759acef548"} +{"id": "train_43378", "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: 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", "governance", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "43f8736406b2352a7d507f9ed14af5d7dd34bc01"} +{"id": "train_43379", "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: 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", "auditability", "tests_are_truth", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "042bacd56eafda5c6d56b38f1375474b76dd3259"} +{"id": "train_43380", "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: 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", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8d0a06454bcb4aae42b9c6fb51af60297377fde"} +{"id": "train_43381", "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: 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", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d504d7474dfa4ca7343e2d1ced550caa9647c4b6"} +{"id": "train_43382", "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: 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", "repo_scale_reasoning", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1dc224985ae95a4d6b84debabc4946561baa0787"} +{"id": "train_43383", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39be158d937c85fe9edf836a52dac27cf16e962e"} +{"id": "train_43384", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d338a6c6dec72ea224c275a4aa22522421759f0b"} +{"id": "train_43385", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "tests_are_truth", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8804f586605f7ce8ff06c20c3c68502783ca8bb1"} +{"id": "train_43386", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55afc3f60d169d2af6600f468b9aecd6ae936787"} +{"id": "train_43387", "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: 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", "ci_integration", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af9b7a7f3a45e896b966af5b370bd690f7150b26"} +{"id": "train_43388", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a483e6afdecdb4de5934f43ceb68749eeec3350"} +{"id": "train_43389", "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: 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": ["evaluation_metrics", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f861d7f64290444053ba8e77df8c4cd2ada7d6d"} +{"id": "train_43390", "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": ["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", "security_gates", "reproducibility"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20f7d76152a2ea5b5b7af37d3ff9c9de619653fa"} +{"id": "train_43391", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f3e6b570d6ab12d12cbb9317b75df6421805f51"} +{"id": "train_43392", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9920232e4867bc256261860e2b64d6f8a0fd0275"} +{"id": "train_43393", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9fc4fe88ad76094a3c0f0c0be2fff21e23a0eaf"} +{"id": "train_43394", "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": ["reproducibility", "cost_latency_tradeoffs", "evaluation_metrics", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6614c1fbe3196ee6e8719619be3b71919cac58b"} +{"id": "train_43395", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "governance", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "885366d1dec0acf42177e53ac3dafc30f60d31a3"} +{"id": "train_43396", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa377164f550c87214e5cea8a7f66d0e7c2d42b8"} +{"id": "train_43397", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "security_gates", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57c724e4e5bcc0b3de2c7e0b6eab5f7b82fbc24b"} +{"id": "train_43398", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["tooling", "reproducibility", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef980fffcbb60bfc2310e9b2e18ef8e838b8989f"} +{"id": "train_43399", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c812d84a5b364b4135a5a23a28a07c123a078c2"} +{"id": "train_43400", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95bdcdeaff5ea53803f16b12b1a6c1b11d122292"} +{"id": "train_43401", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af7056dfba09292316178d19999a152b90edd745"} +{"id": "train_43402", "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: 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", "auditability", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff40f5e64d74c5302cd04f2768e7a1dcc21961c8"} +{"id": "train_43403", "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: 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", "tooling", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f20435852eeec9baf1c61dc92e68c14a0c10e076"} +{"id": "train_43404", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b32fb3b00e9d43796985a5da90a77cd5607e0984"} +{"id": "train_43405", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87fa38c5d8fadaac930e75868e8bdb9edbb5ef24"} +{"id": "train_43406", "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: 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", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57fa72ad742971faf543830ccf73ed5658171dd6"} +{"id": "train_43407", "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: 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", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bc47f9436d49ec2de872e9456516c1d3bb9cda0"} +{"id": "train_43408", "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: 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": ["reproducibility", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f819b84d1dafff28ba76432b9381fa03c9c03c8"} +{"id": "train_43409", "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: 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", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b176dcceee8e0eef1bd7cfdae708f2b6fc491514"} +{"id": "train_43410", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c37c76294970d79af730826148476d6d982bf0a0"} +{"id": "train_43411", "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: 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", "governance", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65cadb27e0e6223498ba4d5f108e99c39e50b05c"} +{"id": "train_43412", "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: 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", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dde6f5b469778d120b340c488cf34301a03e1b39"} +{"id": "train_43413", "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": ["security_gates", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f04b3160916a98ee9f09f7f08bda8645cfa022db"} +{"id": "train_43414", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62d15cfe603cec62790d992f15f83ad853bcaff1"} +{"id": "train_43415", "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: 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", "ci_integration", "documentation", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba2c046fb5bea3a38a2cc19193400bd4f176ce4b"} +{"id": "train_43416", "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: 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": ["reproducibility", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f12216e5ceac981344d390591216c8d59cda9e66"} +{"id": "train_43417", "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": ["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", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f2f36e3dee24967b0766a84ed77d50b91164b88"} +{"id": "train_43418", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4dba916cab25abffca31e3cd99cbf5f6cd2361c"} +{"id": "train_43419", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "779abec667855149c4e5668621f5ea5629e43138"} +{"id": "train_43420", "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: 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", "governance", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c076332bc609741e1181d12fad2650d6934201b"} +{"id": "train_43421", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d3b6c1d9a468552f70beae90976c4a4b688472a"} +{"id": "train_43422", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9124e08a0b3d748bd1de620c45ac4bf5970e98bd"} +{"id": "train_43423", "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: 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": ["governance", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f5a1a9ba316cc36f22aa7eb7f4bc793d000d31c"} +{"id": "train_43424", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "auditability", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d05f40f0beab816ee38ae6118bdb8ed82c99ee8a"} +{"id": "train_43425", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc064a8f7a00c3cea1c320d11b556b799942f720"} +{"id": "train_43426", "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: 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", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be38e62904d376612fe0929ed3e72ebc1c132a45"} +{"id": "train_43427", "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": ["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", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca230a0fc821b9d2f3bc08eddb71ba92926d05e6"} +{"id": "train_43428", "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": ["evaluation_metrics", "reproducibility", "tooling", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fba2b8bc2dc7ebb8790ffe116a70406407430e99"} +{"id": "train_43429", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71d3d247c50c9d6e99450d3136245c1347eabe41"} +{"id": "train_43430", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6757200439f8f444c86e534cb64b75ff62e0c358"} +{"id": "train_43431", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "083550b63f00453999740dc3ed4a5bb23e99fc0c"} +{"id": "train_43432", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5c799b64bb68c0092bbc2c0c1b79e958b2732cf"} +{"id": "train_43433", "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: 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": ["security_gates", "documentation", "ci_integration", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9bf11fc90c7963eb53e14b7d22eada874e09fdc"} +{"id": "train_43434", "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: 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": ["cost_latency_tradeoffs", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e488b1116809634cd45f6e1cb41be2a2cc1541a4"} +{"id": "train_43435", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "documentation", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "885366d1dec0acf42177e53ac3dafc30f60d31a3"} +{"id": "train_43436", "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": ["reproducibility", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c05ef12fe633dec866a5fcdeac40f8d92cf5c0e"} +{"id": "train_43437", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "ci_integration", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c275adc0c5e6747b88dad1e41bd60f22b01560a"} +{"id": "train_43438", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "ci_integration", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e873889b97c2cc4128173c52c7a33ef4c9a77052"} +{"id": "train_43439", "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": ["reproducibility", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74a929bbcc53b8ffba75e05c074ec4d56f65bc10"} +{"id": "train_43440", "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: 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": ["documentation", "ci_integration", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26a7a31543249f832efa2d875a1b57431a2b1e8a"} +{"id": "train_43441", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb6b7ea46cd50c94c463f9a7cedc8d3fe7d38df4"} +{"id": "train_43442", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fa2ce7c1e6552cdbf2e793d916dc12e66e44518"} +{"id": "train_43443", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1151f5bfb9db56112c754c278b0ea1ce84cdc3da"} +{"id": "train_43444", "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": ["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", "documentation", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b3b39d29d50323687f31cdd02ba47dc558580b9"} +{"id": "train_43445", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["auditability", "ci_integration", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdb1ab99a69d5669befe64b2ab91cdf088aad368"} +{"id": "train_43446", "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": ["documentation", "reproducibility", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4fd83a756d23be0ea909dd7efe40f3af4abaa8f"} +{"id": "train_43447", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["documentation", "auditability", "security_gates", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a25229f17f4afb874eb97fc3940c2a9a06b86683"} +{"id": "train_43448", "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: 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", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca8145fd6cb788e00b9de2ba7961db651aacb310"} +{"id": "train_43449", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4aaa77324f2c6067c9dc3595dd609183ec5e1840"} +{"id": "train_43450", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a84f781f10b60ecf95d4e0b12671eae547bb8ab1"} +{"id": "train_43451", "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: 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", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aaf1293581a52633b7c2a147a759b37058848abf"} +{"id": "train_43452", "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": ["reproducibility", "governance", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdc173b23a1b83f706773cfe3a9dfe82d9684ec5"} +{"id": "train_43453", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81fce9a22932d3a1a3f5c4cd4b60d4e2d641aaaa"} +{"id": "train_43454", "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": ["auditability", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b310f11f7f3ef10390e500a8cf7fa903ede3377f"} +{"id": "train_43455", "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: 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", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a81137320f9de1f9c441af58c9e68d6186a698e1"} +{"id": "train_43456", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a05cc3f55686f1ea2b4703eb5fca08c33833b9a"} +{"id": "train_43457", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56d3659565d80fdc36e056f30b05a59872517988"} +{"id": "train_43458", "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: 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", "tooling", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd660703052e00afb303b89aafdccc5550e57fe3"} +{"id": "train_43459", "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": ["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", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b304abf5822694d96f454fd48d7fc490e64ec98"} +{"id": "train_43460", "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": ["documentation", "auditability", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcc0adc0e872f933044f7f4c96b9eddff947815a"} +{"id": "train_43461", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ad53f43e2a165c3ec670ca55cab86399341d6ef"} +{"id": "train_43462", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "auditability", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe9bedbb7ef0798b05d5159f84fff41fca0d0798"} +{"id": "train_43463", "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: 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", "auditability", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "675bb14e778c398f1724bbcd76b8ddf5890ab495"} +{"id": "train_43464", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b658a01f35ab93c6756adda5bcfb5ae2bbf1f34f"} +{"id": "train_43465", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "governance", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87ad400fb436cce7274384967d6e1714d4b9c339"} +{"id": "train_43466", "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: 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": ["repo_scale_reasoning", "ci_integration", "documentation", "governance"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e58a234b9eda94cc4f68edbf0224e8f1339504d9"} +{"id": "train_43467", "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: 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": ["repo_scale_reasoning", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2e8581f2b1868a093af3a408d75cf46224ab6f9"} +{"id": "train_43468", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f52f64a3c728c413c397ecf81446ed7f7b8c929"} +{"id": "train_43469", "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: 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", "governance", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "30eda413a1a6ebcd895031c1c8c54081a67ef409"} +{"id": "train_43470", "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": ["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", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e0afc096d69b65b59c08cbf00ccc39511b049e7"} +{"id": "train_43471", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d5ccceb507d235fe360482bcf77baeaa0298c86"} +{"id": "train_43472", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "repo_scale_reasoning", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "580a39b7191b2df3db058658581bf7336bc18ef4"} +{"id": "train_43473", "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": ["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", "security_gates", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28b8f548ef906092b42b7273994a39fcd901c6aa"} +{"id": "train_43474", "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": ["tooling", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b928f7171c524f2097743553f79f6898b6c68612"} +{"id": "train_43475", "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: 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", "auditability", "governance", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33d6a0afde0bbf720bb5bd97ae00fb4f45787248"} +{"id": "train_43476", "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": ["security_gates", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "96da1556bdc4f2a234c4a1c2d0beb5595c71112c"} +{"id": "train_43477", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c8f4e4aa0911dd7e269271d12e7550b3cb558f6"} +{"id": "train_43478", "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": ["documentation", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be5383a2f1d51c9d89a05243ff2707154859e8c5"} +{"id": "train_43479", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45c16333a981f67f0518c90df44c0f6a0f0d9cde"} +{"id": "train_43480", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "ci_integration", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ced3f3f71a3741bde8ae23c3ea66bc4e2e50fb2d"} +{"id": "train_43481", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1fd8e46d201ecc2b0228d7c3e9aec1a4f40feb13"} +{"id": "train_43482", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ff176150b2edbc1082403be4038baf05a84fe4d"} +{"id": "train_43483", "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: 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": ["tests_are_truth", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "955779a2b5702b430bbe2a9b1d8af68b85a74b5a"} +{"id": "train_43484", "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: 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", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccc0c03e3a8d0f5d432c310f9d717a46cfe33f7f"} +{"id": "train_43485", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c44fb00b2fd77c2020b7097fcc440a310956456"} +{"id": "train_43486", "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: 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": ["tooling", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "047b51618d801132fbb91b342b8e2aa8b32ee748"} +{"id": "train_43487", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e5307a9bf15eda34502890239eed11f8181731ea"} +{"id": "train_43488", "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: 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": ["reproducibility", "ci_integration", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf0349e39214b50118aa581597e83e0a2f03de04"} +{"id": "train_43489", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fd159152d9005b16ad4405b762570fd5241fb84"} +{"id": "train_43490", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "reproducibility", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daa6fc42a7b7303c918f616bbd3ade7841b99b63"} +{"id": "train_43491", "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": ["security_gates", "cost_latency_tradeoffs", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8a47e82babe59f13035a119e38591fda61171dd"} +{"id": "train_43492", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc07fe33f05cf1b570cc09ee1d3a057a0747fad7"} +{"id": "train_43493", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "642e53407a606990b946c5f910e6f862c4fb5a5c"} +{"id": "train_43494", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["reproducibility", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ee1e4e784c6a400d44c6852d408d05cc84a0760"} +{"id": "train_43495", "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", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04b4e73c335db953c4c160d380ae3ae7b3a47397"} +{"id": "train_43496", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e325fe81ac2f8389a4b3a65ee4a2cbc2e29c670c"} +{"id": "train_43497", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "tooling", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4dba916cab25abffca31e3cd99cbf5f6cd2361c"} +{"id": "train_43498", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19254a810261c6039c4a6a5371cb9247197ccae3"} +{"id": "train_43499", "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: 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": ["tooling", "documentation", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f004a01f9a01e59effafa556f3ee8967243270d"} +{"id": "train_43500", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa451e077559708cd3d7b9dae0dee3eb7bbfa2a3"} +{"id": "train_43501", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "923fa5adf4b9db4f8925af4ad44a9e2322428c50"} +{"id": "train_43502", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "ci_integration", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec4fecf817d94c20474a85579b77b548b164902d"} +{"id": "train_43503", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb08ec9fbc2e660b779b219ccdef22b1e03d83ff"} +{"id": "train_43504", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2dea667d415a196663816053d2308584beb1f7b2"} +{"id": "train_43505", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3274dcfa40ba487eb383eb650d30207ccb8186b"} +{"id": "train_43506", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aadbab002d1793daad9e94275ab5368c2cdebf9f"} +{"id": "train_43507", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "security_gates", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ce51f9cae436ca38998375dc96cc47fffc5cf16"} +{"id": "train_43508", "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": ["security_gates", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b50574c26689ca0e02bc26cf2b5f94af3cdd3c60"} +{"id": "train_43509", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5cb906f5bd7e403094989d058d58b1e956381df2"} +{"id": "train_43510", "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: 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": ["documentation", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "731de9d90ac2b151f2081e0085fce100b218874a"} +{"id": "train_43511", "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": ["Identify failing behavior via tests", "Minimize change surface", "Apply patch", "Run targeted + full suite", "Verify no regressions"], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc13d560a57145418916f5014c90631f2caee234"} +{"id": "train_43512", "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: 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": ["reproducibility", "tests_are_truth", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c660cfdbafe0627f159ed26805151843aa38027"} +{"id": "train_43513", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bc3dc832569a02d782f50a462648c09cf80db2c"} +{"id": "train_43514", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "850c9df216639189b12d76a62c595b0b7cbd68c0"} +{"id": "train_43515", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53e1d01d7d9396b05698f26035c6ab3c9584772d"} +{"id": "train_43516", "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: 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": ["auditability", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9382ff9a7e0163311f59f369472e9ba014750e98"} +{"id": "train_43517", "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: 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", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a5cb8c0c7b6152679f37fa7c6f5c8a26e968001"} +{"id": "train_43518", "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: 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", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "704bb005731101e7369696ba87f47553af503f7e"} +{"id": "train_43519", "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": ["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", "ci_integration", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9ba0aa6425babcaefcc0ba030f0b5901b492653"} +{"id": "train_43520", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13a6cb0a7a8ba5e0d6cae9d2675a99d621978212"} +{"id": "train_43521", "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: 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", "documentation", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e451a4f3b56baa760b52131d437b6a1bc05bbe36"} +{"id": "train_43522", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "030fa9bf66e77a51937ab2c7726060e5c35d8eea"} +{"id": "train_43523", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b9ace2188694902cef786f933c6df2af9a6ae6c"} +{"id": "train_43524", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "288bb1b86c4cb1b7c6e0b8c39103b2309fbd6385"} +{"id": "train_43525", "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": ["reproducibility", "evaluation_metrics", "governance", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d87844f09f7425293ecbf4b3d7332df049c4d1a8"} +{"id": "train_43526", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["security_gates", "tests_are_truth", "repo_scale_reasoning", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4532bcd9c8fa71ca73d31fed3ba4fa85403f9025"} +{"id": "train_43527", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f8eaf8a31f9d13c643c31b27152c560dae3297d"} +{"id": "train_43528", "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: 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": ["ci_integration", "reproducibility", "tooling", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3deb75971ce256de4d3a0fe9d02c673c662bc6f8"} +{"id": "train_43529", "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: 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": ["tooling", "ci_integration", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e88002a3b1f4d21df368709dffc2cd88adff3d7"} +{"id": "train_43530", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cff4442441793cad7a85925fb19482d2e25ff625"} +{"id": "train_43531", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4b9cc15e711b3dfff61819045bf179d5c8ecf6ad"} +{"id": "train_43532", "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: 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": ["documentation", "reproducibility", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efef7672cf8a6498ec161ab57ff0c79b8ed87f09"} +{"id": "train_43533", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ba06ddb6e5af1a0a482f09790ffc87c606907a8"} +{"id": "train_43534", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05adc8e9d82f99f981fb7886f20f969313dd3c98"} +{"id": "train_43535", "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": ["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", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0978f0bef8473adab6a2636fbc719148bad6b960"} +{"id": "train_43536", "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: 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", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d2e8581f2b1868a093af3a408d75cf46224ab6f9"} +{"id": "train_43537", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1bcd784b9dba6574ae341de9e0a38b6bd87de2fb"} +{"id": "train_43538", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a62705be509bab230b406e90a4cfa81ad2492749"} +{"id": "train_43539", "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": ["tooling", "ci_integration", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "542bd5144fbfb125b86d3123a4951b3715b6ab79"} +{"id": "train_43540", "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: 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", "cost_latency_tradeoffs", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e349a0114c3074380a490a98385a35cd19169bae"} +{"id": "train_43541", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08e7cabe873c186c049a347e22d048a4cc7ba31d"} +{"id": "train_43542", "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: 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": ["auditability", "evaluation_metrics", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afecf56309acd34a4bb91cad96abffbab8e31d42"} +{"id": "train_43543", "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: 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", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8ba73dc999216e6829740e605d30c4592349d45"} +{"id": "train_43544", "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: 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": ["ci_integration", "tests_are_truth", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f43e1d43e980b6fc7776f50bc1c7b1390ee525e5"} +{"id": "train_43545", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8da6080d73f1970c545a5968b4f6663c3b09c8df"} +{"id": "train_43546", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4425c4f41edf4acc2feb0f69d999a7847f1dfaaf"} +{"id": "train_43547", "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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae8a40f6e80f927f71850df45e1c105541a3a2cf"} +{"id": "train_43548", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be15e2cc4b72ffc14ab16d872a2fdd1475796067"} +{"id": "train_43549", "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: 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": ["documentation", "auditability", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6aa992ccd0d5c4cf7a3898367797d533fe2d116"} +{"id": "train_43550", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a8e3bdf1876e9dfabd42732701a473bc4db9954"} +{"id": "train_43551", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "ci_integration", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9c8d78ca0ba342f51e60294f2e97469b5d109f0"} +{"id": "train_43552", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8417e5bb771e346e274e1739379c88de5194b16d"} +{"id": "train_43553", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35679ecfcbc8bba3c2e35735e5636ca99145fd35"} +{"id": "train_43554", "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": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8a47e82babe59f13035a119e38591fda61171dd"} +{"id": "train_43555", "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: 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": ["reproducibility", "documentation", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13381b03849540884ab727b6d34f4ef97d541550"} +{"id": "train_43556", "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": ["governance", "documentation", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bc5725ef56568a3f6d0cc391a36cb0a3e8bc145"} +{"id": "train_43557", "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": ["governance", "ci_integration", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71aa36d01fd476d061e65a302346673a07f3495a"} +{"id": "train_43558", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["documentation", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "443ec1c4e0fac27110083e4162747d7c4e51aa97"} +{"id": "train_43559", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "ci_integration", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed1b12ab2e3fbb789ba974df04e12f3d8813cd71"} +{"id": "train_43560", "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: 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", "auditability", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a2dbcf43c1db30c7bc6db3238744a66c31e7093"} +{"id": "train_43561", "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: 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", "governance", "auditability", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3df1fa9cf9e391f9e855210c66d67066ca60ed95"} +{"id": "train_43562", "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: 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": ["security_gates", "cost_latency_tradeoffs", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b2b83e389875913a29db6e38ccef26e13a205ff"} +{"id": "train_43563", "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": ["auditability", "evaluation_metrics", "security_gates", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f104700d81732069f945b8a9e01f460e8818b609"} +{"id": "train_43564", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "evaluation_metrics", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dcc10d39a5d8a35f1655ffa1ece69b0dd0304bba"} +{"id": "train_43565", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbbd2be0c5149beff84b8ed2140a18bf95b19b4e"} +{"id": "train_43566", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["auditability", "evaluation_metrics", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9466ddf7c0b3e5e7fb0ecf536cbe5e6676f4f0c9"} +{"id": "train_43567", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["reproducibility", "cost_latency_tradeoffs", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d34b60b04ac2f65d17935f058bfce4c39c5fbda"} +{"id": "train_43568", "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": ["reproducibility", "auditability", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f0c628a7240fb99ae81c8740fe4eb1910e93c14"} +{"id": "train_43569", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cb43cf420be6f99797025a71142bf086abdbf13"} +{"id": "train_43570", "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", "tests_are_truth", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ef41f2ce9da4a95da6f24ea02c94ac6f3178df2"} +{"id": "train_43571", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "463ef235f949682101474b1f40ab45f152a44e66"} +{"id": "train_43572", "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": ["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", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "598423a662eaa5e7f0b214e4d701ab5823f97c3e"} +{"id": "train_43573", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34b3b9570caf83ce03a48c110f4562a1fd498839"} +{"id": "train_43574", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0aa634b478ff648561a5efc7fa3d89be0bd20579"} +{"id": "train_43575", "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: 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", "tooling", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "862342d4a76053be1942621027d0204c6680571a"} +{"id": "train_43576", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability", "security_gates", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f66268f32345c71b216b447ebd03e73851908776"} +{"id": "train_43577", "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: 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", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d4b663604e7f54b59f477f7baf8e87474aafe83"} +{"id": "train_43578", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdeccd78d689710d4966dd856253f46a4e42a58c"} +{"id": "train_43579", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["evaluation_metrics", "reproducibility", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f22fdb91d9aca449debd540770e3d69ad91332e4"} +{"id": "train_43580", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0466cffbdf3a4ef8adb4109bd043d6ab9095b776"} +{"id": "train_43581", "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: 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": ["auditability", "repo_scale_reasoning", "governance", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41f29ddd86a3147e3022b14f328be81de744945a"} +{"id": "train_43582", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eec21ef77a13e8ad8705691b86a158efcf955eb9"} +{"id": "train_43583", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25171d09929af7ad2b02c16b5d7813eadbabe616"} +{"id": "train_43584", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "075e3e09737b09e6fd4d076111d954d29938f5d4"} +{"id": "train_43585", "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: 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", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b59d1bc448e936582d93e2a782a68cdbc4e3068"} +{"id": "train_43586", "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": ["tooling", "reproducibility", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc2d5a3c63c6f5ad0af2edf3269a5282ff6a5800"} +{"id": "train_43587", "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: 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", "documentation", "governance", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36b50ad52aed500e30f096e5d06cfdd210eeb0c0"} +{"id": "train_43588", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2ccd8a0d1ede5d121e9ef8c91661ef2f28d82d7"} +{"id": "train_43589", "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": ["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", "ci_integration", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3e3a87ab9d3593fedb12c197be5de81ee8b2fbe"} +{"id": "train_43590", "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: 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", "governance", "documentation", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc171872647a540aa3dec5f515f9887ec5a6bd58"} +{"id": "train_43591", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready 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", "governance", "documentation", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "498cb4f94e4f0d196667c4452f5cdafc1fc98ec2"} +{"id": "train_43592", "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: 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", "governance", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17f53a50b9bbf4dfd64d7ffda5411615e99e2e1e"} +{"id": "train_43593", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f75796afbd69ad839a014586c13401988eb59122"} +{"id": "train_43594", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "documentation", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a390e646bc947ffd7d609ed3525f01581e6394b"} +{"id": "train_43595", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "documentation", "auditability", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e23989027bb01cdcb89620b13ca2ad61b850cd31"} +{"id": "train_43596", "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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "99542cca70308c9466dc68c7c2a2165295f93cef"} +{"id": "train_43597", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0176bdcec4b5008b2416d5625382a994f06b9ccd"} +{"id": "train_43598", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20f17cff070e3b1f9419894c2c62b6163ebe050d"} +{"id": "train_43599", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "tooling", "auditability", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "214313c47b66218f466ec151cd7714113228181f"} +{"id": "train_43600", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b6e450ac9253482ceebc16f94b0fa810c602bd3"} +{"id": "train_43601", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["ci_integration", "security_gates", "tooling", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6622daca6567d2a3dd54ac54ae56000f25334493"} +{"id": "train_43602", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "governance", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8174a8343da5405641307491fe80674ba559280"} +{"id": "train_43603", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["security_gates", "auditability", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9571e5c91be2d7ab32551c63fd251e48639e10d"} +{"id": "train_43604", "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: 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", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "394f07a7c3964d293bde17d8a9b9dbdc397acfcb"} +{"id": "train_43605", "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": ["documentation", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b36edfc37f0408af561504af876afdcf45f59c2"} +{"id": "train_43606", "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": ["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", "tooling", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e691464c2582655d26f1db4e376cbf8722ad7b6b"} +{"id": "train_43607", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "179bc2a19c75e8a36b293f43c0b40a4d4d7a6364"} +{"id": "train_43608", "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": ["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", "tests_are_truth", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0597e3c7c40b7b6795f579f6e5f4ea7f3d7bf672"} +{"id": "train_43609", "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: 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": ["evaluation_metrics", "governance", "tooling", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1008a779e243f431f39b46136c201c07744693d4"} +{"id": "train_43610", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9aa0a7a75707a078534560c5cb3a28646163d2f"} +{"id": "train_43611", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c6d3d30a7d397ad608afd43b662e5a5b332d3018"} +{"id": "train_43612", "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: 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", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "272b16df03c91290b9bed7ea5a64ad7f9cb983e6"} +{"id": "train_43613", "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": ["security_gates", "tooling", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91b22f68ea9214948e39b262402a5fd4cb89458b"} +{"id": "train_43614", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["documentation", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a214d75b27326e37a45e963669b6ddfe43d2c72"} +{"id": "train_43615", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25c044c4b7412ae57af73a9bd74653ba918ef589"} +{"id": "train_43616", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e72d76c8188fd3135b23d89f658c80e54827520"} +{"id": "train_43617", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb12c9c359f93f3b1364b931f37f4f4ec5710a7f"} +{"id": "train_43618", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f55d13cbc904691c51ece1bc78d41114c936545"} +{"id": "train_43619", "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: 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", "cost_latency_tradeoffs", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f83524357083e442397b2985e3b3e3657b19ea65"} +{"id": "train_43620", "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: 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": ["ci_integration", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5d3ad2c54528e9294f37fd0ed1d596f272896fe"} +{"id": "train_43621", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38ab4f9d6838203c5de30a08fe8b8afb40c6f744"} +{"id": "train_43622", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "governance", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a32a26cd5d71fa7421068aa662fbfdcb6dd5ef88"} +{"id": "train_43623", "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": ["tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd1fc3b2405e8204299b1a368af34332d39fa716"} +{"id": "train_43624", "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: 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", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "230a0a9c311863f25edc1d841cad71e51eeffc74"} +{"id": "train_43625", "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": ["governance", "tooling", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "679044c575fbcb6e4b462cb5919d4230cd61153a"} +{"id": "train_43626", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82719cc96e389770ac27ea054a22ee5d9e73774d"} +{"id": "train_43627", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88a3c227e288e49dd6f60192f61b67b9075ec424"} +{"id": "train_43628", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9884a5c6cec7417ec93d9cdd7c56ddc608829fad"} +{"id": "train_43629", "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: 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": ["cost_latency_tradeoffs", "security_gates", "tooling", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd47d038250a73f878692e4538243f2e96708ef3"} +{"id": "train_43630", "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": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62f0bcad36891d67012cea4a580bb2f157419ca0"} +{"id": "train_43631", "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": ["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", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "548957317b48e1e6b8beaecdc438f4e7f154f63a"} +{"id": "train_43632", "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": ["governance", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "429020ec1bc9a558ea75c6f4fa49ae5dee2fff3a"} +{"id": "train_43633", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["tests_are_truth", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78a9795e2c537c11f534a8ecf2d0666ff765f509"} +{"id": "train_43634", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "reproducibility", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02bbd351d685048a1e645d3f8dcc28e10cb8e18c"} +{"id": "train_43635", "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": ["security_gates", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0313224fe3bdffc5bf6a6ab5ef0e384b7d52df52"} +{"id": "train_43636", "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: 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": ["repo_scale_reasoning", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac6fa8a6658274020c531d4ae811a3246964b6d3"} +{"id": "train_43637", "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": ["governance", "evaluation_metrics", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "680dcc25f2cb7ed6307bf4b8400008932a12fd2a"} +{"id": "train_43638", "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: 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", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c753751f35eca02b90c5e4c60b015184deb15fca"} +{"id": "train_43639", "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: 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": ["tooling", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98b94e87e585482e8807da7b3e6a565672d6c7cf"} +{"id": "train_43640", "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: 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", "documentation", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ade457347bf1f1f5a7994c5523cbcf29b5c32b85"} +{"id": "train_43641", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ba5cbeac014a018f280276b48d74c4911e0f51b"} +{"id": "train_43642", "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: 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": ["repo_scale_reasoning", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a188b3190601a08d5fecba68964d3245567ca48"} +{"id": "train_43643", "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: 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": ["auditability", "tooling", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bde30b3a9205c2e9f09f0d1c98e33e51def204dd"} +{"id": "train_43644", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7dd00fce40f8dd36e9d60556df393056291f903"} +{"id": "train_43645", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5f950bf499cc77e3efb651ea185ffadf6429eeb"} +{"id": "train_43646", "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: 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": ["governance", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5710ed505f6896859432c7aa6adaa6704a91c0a"} +{"id": "train_43647", "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: 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": ["evaluation_metrics", "tooling", "documentation", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e82f016b2c93351b5d76b0416384e8692ce13211"} +{"id": "train_43648", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["repo_scale_reasoning", "security_gates", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "817e32321ddda7533f9e1ecc16d9f6b0b53a5e3a"} +{"id": "train_43649", "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": ["auditability", "documentation", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af0a1ec68f5fb69c31aa0dd02f01824b6a810a07"} +{"id": "train_43650", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["reproducibility", "ci_integration", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3ed126e10e6a63a3e727fcba62a278b8d53b96e"} +{"id": "train_43651", "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: 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8ba73dc999216e6829740e605d30c4592349d45"} +{"id": "train_43652", "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: 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": ["security_gates", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "395b0c8b4ef3442d0ec56706df9288057365e5ed"} +{"id": "train_43653", "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: 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": ["cost_latency_tradeoffs", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf0991058eef10b482cae76b7ebca2bddeb7d69c"} +{"id": "train_43654", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87a5263c3f1344c6a407aabe57d92c15c43a8f03"} +{"id": "train_43655", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["security_gates", "ci_integration", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00b66deecb488eb8c3df69663740c54d9e0fb2c0"} +{"id": "train_43656", "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: 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", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e11a9309c3ca2575499f87a7931989ed28f2946"} +{"id": "train_43657", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "267920ee4bfc6786cb7588bbd588b2947046afea"} +{"id": "train_43658", "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", "reproducibility", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c88eb69f0f77bdcd1a2e4c5482dc18bb3fc492b8"} +{"id": "train_43659", "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: 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", "documentation", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "567c93fa0d29aeba13fcd897e039e8307188d85a"} +{"id": "train_43660", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88c8f7271777bcb5552929358eb4211c8715de03"} +{"id": "train_43661", "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: 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1ff4289366ab49b380e5f80de71ef692911cb15"} +{"id": "train_43662", "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": ["ci_integration", "security_gates", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf4e71d4f16493e21266dd9f46d67d46cb02e6ba"} +{"id": "train_43663", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "181654a103c17503eebf871c648637e5458feb7f"} +{"id": "train_43664", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9a81afb2c3b3ac24e76a6747c13d7a33c594ca1"} +{"id": "train_43665", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcc2da3e8d5427800516adb732de3335e6c9cba1"} +{"id": "train_43666", "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: 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": ["reproducibility", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9dc875e337e4aa3bcc127840e25b7f8a7a5b5ed3"} +{"id": "train_43667", "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: 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", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5cfa908d17cf4eb3f5c3b862eacd56dd10729fd8"} +{"id": "train_43668", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e487bd7bac0b934bf5df5e7c6817d34e30e5cf4f"} +{"id": "train_43669", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "reproducibility", "security_gates", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5da81fdb6dbd0b29cc75f1b9f28143b06629ba9a"} +{"id": "train_43670", "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: 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", "evaluation_metrics", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb4953a7bb1cc6b72a155babc642bfff25b16fdc"} +{"id": "train_43671", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ba9e42bb469842b6ef61b2979c81fa1b5c7f3d8"} +{"id": "train_43672", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4379adf174aaa657dc98fcf734d986dccdef72da"} +{"id": "train_43673", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e49735f10d3ba1809dc1a4bb02915bdd896a911"} +{"id": "train_43674", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85763caac51b7d2b044b0668d8158dcd86384e07"} +{"id": "train_43675", "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: 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": ["auditability", "tests_are_truth", "security_gates", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c15fe80f7757d997c7462a5237b1e97386dd19f3"} +{"id": "train_43676", "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: 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", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2aeccc75376e3229bdc026ca9d723919e5cef628"} +{"id": "train_43677", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ea4e650e998be5ca1a1793480e70a25710a94f8"} +{"id": "train_43678", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ca09ccfc81ecef00a1777e39d1e7c48f6d7cd31"} +{"id": "train_43679", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "895265a23e1afc8fe88eab96be77eb0cc32d3e34"} +{"id": "train_43680", "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": ["governance", "auditability", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da49f2ef73a859a84b182d697d2a32d185786253"} +{"id": "train_43681", "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: 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": ["evaluation_metrics", "documentation", "reproducibility", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56f5330b09d33f479605f27ea0c3a9b0112686b3"} +{"id": "train_43682", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9762443951d03bb05c2ce971e259b7d96bf59d05"} +{"id": "train_43683", "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: 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": ["tests_are_truth", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b6ff48032b31eb606500509d562e5be077674df"} +{"id": "train_43684", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bb6546dab3ad8f5d0e49afe9126699167f0e628"} +{"id": "train_43685", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["documentation", "tooling", "evaluation_metrics", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "709e124b9668e18d572118938efb6c852abe28c8"} +{"id": "train_43686", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7accb4d6051e4ec1a92f8b97956351786ddb174"} +{"id": "train_43687", "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": ["tooling", "documentation", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b214b97fca46a3628cf3847e3d2f1ffbd147f5c"} +{"id": "train_43688", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "reproducibility", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a390e646bc947ffd7d609ed3525f01581e6394b"} +{"id": "train_43689", "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": ["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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd9fd69273a26ef37e26d2b222c7c440401d64a6"} +{"id": "train_43690", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8b6e450ac9253482ceebc16f94b0fa810c602bd3"} +{"id": "train_43691", "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: 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", "governance", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ee158e9bb24006bc9464ebd6d7357c71612aa63c"} +{"id": "train_43692", "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": ["tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9fc24e59f14b9f209c8adc77933fb8a945e2773"} +{"id": "train_43693", "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: 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", "governance", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c478cbcd274a1a2633b448487a749789426fa9f"} +{"id": "train_43694", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "35267c1a0c4f564d4d8977a0ee0678d299168c1e"} +{"id": "train_43695", "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: 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", "ci_integration", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2629ca7f2a1bf9ebe1d0116f5d5e5cfe3cffa534"} +{"id": "train_43696", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19b21cbc017e06971677ca915c4a50e8e5aef0cc"} +{"id": "train_43697", "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": ["tooling", "governance", "reproducibility", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c52aaf9b1dc38c64eca808502da6cebdb23933a6"} +{"id": "train_43698", "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: 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", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9aa7b0c2494f9fff6899a8237461a9a6994a89e4"} +{"id": "train_43699", "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": ["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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fba9d0f0ba1e6c08baf7b6c28661fc2a8dee96d5"} +{"id": "train_43700", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de04651a4e8a77b5b319e38dca11d727f8244b69"} +{"id": "train_43701", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7703a4b1c92b82ee9a76df754ab18a08266ec365"} +{"id": "train_43702", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fe3bc5d9d6199f06f6ff0eeb40d25d3f0504100"} +{"id": "train_43703", "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": ["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", "security_gates", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6937b78c23dfc6d52c3550e37c6cd59200224028"} +{"id": "train_43704", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a72187bef4a32e9da7b609593101ce7719f3275"} +{"id": "train_43705", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f847235f55a1f8af2b755684037156d18f65be03"} +{"id": "train_43706", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eebeb89bfaa6bfb01f6956937e2d8bb167bcd63a"} +{"id": "train_43707", "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": ["ci_integration", "governance", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0708930087b4e77ae2439ad3cc469b4a227b8e09"} +{"id": "train_43708", "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: 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": ["governance", "auditability", "tooling", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4771fa14b370eef8bdc3d3300c3afae49476b65"} +{"id": "train_43709", "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: 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": ["auditability", "repo_scale_reasoning", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e350455146e758bb8b9c2ee38f92a7e5a3c607b"} +{"id": "train_43710", "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": ["tests_are_truth", "evaluation_metrics", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5869ebbfbab195867ea70300fd8bfa4080217117"} +{"id": "train_43711", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1be04c48c8332cda3d12713721cdbb2ca3bb05bc"} +{"id": "train_43712", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea7cdcefcc92ff9dd522a77ef6fa67505da6e373"} +{"id": "train_43713", "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: 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "761c61e69da50b12c854c02304b80a3f6b2e1b42"} +{"id": "train_43714", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "beb25849ed0fb5ad2041b7d9f667371c292d3194"} +{"id": "train_43715", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "110892f60e28624c5cfd40a7832be3b585662a58"} +{"id": "train_43716", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69cc06b49c8c54a9f1ed35bfe38fb46e4b4b1129"} +{"id": "train_43717", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac7e7968380a71de8faa546650e5b9d86b5967be"} +{"id": "train_43718", "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": ["cost_latency_tradeoffs", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac7c94740920cd2959f435945b7f13ff1d0c804a"} +{"id": "train_43719", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04641b0cd7759bf9611e6ec073a9774678f37e06"} +{"id": "train_43720", "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: 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", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5041fe04d890dcb7ca5a7ffe24fbd4a6a59699d"} +{"id": "train_43721", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "329dbf431524f8bd326945391b1c74119d464ea4"} +{"id": "train_43722", "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: 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": ["auditability", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "142d7f60d1fdca4fe30010e78069e6b39ee21cf7"} +{"id": "train_43723", "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: 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", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4b9b8b499766f44f40429df69902de452555b6a"} +{"id": "train_43724", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dbb10ec3d3d9386eb8d55421921581290434c81c"} +{"id": "train_43725", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eea461906be234e1449eeb286a903268fa98c6ea"} +{"id": "train_43726", "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: 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", "governance", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ba60deba972a2dedf60b2a68190cf41bd4327ad"} +{"id": "train_43727", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17dbd1e5d118c0dda985909f156919afbfdc0f5d"} +{"id": "train_43728", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d66c69c4508b3b737cd320b6b2e649b85214005"} +{"id": "train_43729", "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: 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": ["documentation", "repo_scale_reasoning", "security_gates", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ee9b87cdb452bacabb7ea948c2da483b8c6f6e0"} +{"id": "train_43730", "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": ["ci_integration", "auditability", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31119fc39547d5d1ffe9b50fecd13dad057b9352"} +{"id": "train_43731", "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: 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": ["tests_are_truth", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69dba35c2384177e4ccf7abccb57c5c07b326dfe"} +{"id": "train_43732", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "auditability", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2400b742dbdee57a2fb4c4be64ce9e14f0ecaf9"} +{"id": "train_43733", "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": ["cost_latency_tradeoffs", "ci_integration", "governance", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c47ecb6f13ac3844e3737e556ab0ed66d042fb8"} +{"id": "train_43734", "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: 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": ["governance", "ci_integration", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5360532ca720beb24781bd2b6885f6361849033e"} +{"id": "train_43735", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e14adf8722303ff71ad8ef90aeb91588d973c90f"} +{"id": "train_43736", "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: 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": ["auditability", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5a7085bbd8e0063738302792f2b0d6b83e9a50c"} +{"id": "train_43737", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca79cddeb3de1d31e8e660b57f9ae559018216af"} +{"id": "train_43738", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3908fd3e2d068d8b7416acbad8fbb808641585fa"} +{"id": "train_43739", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tests_are_truth", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5afa86e27d0699ec1927dd6dc048eb8de77c557b"} +{"id": "train_43740", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d04e9d678d962445c9cef052cd0c58480c6ab1fd"} +{"id": "train_43741", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe98195c38de21b4bb6ac8d3992402c530c65b39"} +{"id": "train_43742", "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: 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": ["tooling", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce47d19b83934b8a3c5496abbf50ae2320412711"} +{"id": "train_43743", "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: 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", "ci_integration", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "764e89c29b808941da076134835c0f097ea23820"} +{"id": "train_43744", "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: 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", "governance", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a81e247249f0ee46319f36b4b4d6d8b153c0e0a"} +{"id": "train_43745", "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: 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", "tests_are_truth", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a01dac3255fe0f380672842133a95f7beed8fdd"} +{"id": "train_43746", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "486aabad79ecfea1f15629b39ae06daa7d119c31"} +{"id": "train_43747", "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": ["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", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fea0bd741622a3dff2cb38856dc1b49232f37001"} +{"id": "train_43748", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "469a14d3a59d0f5d5202893c7a299b87aad54df7"} +{"id": "train_43749", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "security_gates", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3c2d6b113a7bd685a66a3c6916c1f54499f92fc"} +{"id": "train_43750", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "documentation", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ea4e650e998be5ca1a1793480e70a25710a94f8"} +{"id": "train_43751", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "564a212cdee0803c142848366fa777aada2b4151"} +{"id": "train_43752", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d8b1778c7b18bcb109fe931c817ba4321808146"} +{"id": "train_43753", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "documentation", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f621e30bd75e0958c61a42d5a4b8627f8c1219be"} +{"id": "train_43754", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "cost_latency_tradeoffs", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1583da0da539a0f6b04ca291ce1ef0c390f5ffec"} +{"id": "train_43755", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "654f847fd62e18188e00b0abfd2670821808b034"} +{"id": "train_43756", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready 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", "ci_integration", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e93b1be6fbe01e68176f228a57cdcd3fa67d4853"} +{"id": "train_43757", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c55c493b5c529d53d7dceb34ed0f3b0eb015ea79"} +{"id": "train_43758", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cee6cb588d4a3c2dca32401f5ac74dbc624198ff"} +{"id": "train_43759", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "775c46b625bf4e7c7c10ee4ddd9ea29ff14fd437"} +{"id": "train_43760", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cff003dbab5f1aeec44c12c761c270d4fe9e694c"} +{"id": "train_43761", "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: 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": ["tooling", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75adbc8e78274a5f19ff95fad2d764dd7834cc1d"} +{"id": "train_43762", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc41061f1ad8a28d2443b59a9655679d93a7d6cf"} +{"id": "train_43763", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df146e49562dbcd135193a7957623d9ffabb804c"} +{"id": "train_43764", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "security_gates", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "961069b6a8323e2c7ef540a120579c887df74e50"} +{"id": "train_43765", "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: 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": ["evaluation_metrics", "documentation", "security_gates", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "058fe2ec8bcd61067d2f13ec42e5b01ffb9415f1"} +{"id": "train_43766", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2400b742dbdee57a2fb4c4be64ce9e14f0ecaf9"} +{"id": "train_43767", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "244169bebccaabe3d8e94ba33132d8ef6a43569c"} +{"id": "train_43768", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64d74421348daa09cf2fb34d3b36960f6080a9b0"} +{"id": "train_43769", "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": ["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", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2fd1495f515b54ba3598099231f93f697bfa340"} +{"id": "train_43770", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a86c999d732279dec428821b625d14cd47582e0c"} +{"id": "train_43771", "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: 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": ["ci_integration", "auditability", "documentation", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8dbd10b3711f73d65883db8efc794bb8b1299b86"} +{"id": "train_43772", "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: 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", "evaluation_metrics", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ade0c0dcbdc1ee19b9af0e484890f1c3a2399cf"} +{"id": "train_43773", "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: 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", "ci_integration", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a40e166fc07ac07957eef161fd7115cc6369e7f0"} +{"id": "train_43774", "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": ["cost_latency_tradeoffs", "ci_integration", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a98577db8acd1f9e340f15bd41eaac52a3d958a2"} +{"id": "train_43775", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5b47fe764be3d653dedad3ced26910534e7cfd5"} +{"id": "train_43776", "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: 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", "security_gates", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a116631863fbe3fef27965f695203d292691032"} +{"id": "train_43777", "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": ["tooling", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3774de09a4482467a43ee00f34f990ec261f3536"} +{"id": "train_43778", "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": ["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", "tooling", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65dddcded88f699e01ddf12aa6d1157ba5c110a1"} +{"id": "train_43779", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b919603f2b9831f2df289efcf3b2e76e934630e6"} +{"id": "train_43780", "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: 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": ["repo_scale_reasoning", "ci_integration", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "252f8276b29ee6ce38ee7efcd4856a3f66478e10"} +{"id": "train_43781", "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: 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", "tests_are_truth", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb568292982c8f54b8b7f25f82fb9e0b651452a6"} +{"id": "train_43782", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12a2f26e7a32834ffab287b7aca5108f01ab7468"} +{"id": "train_43783", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44713010d07ede19a148958d49af2a8f1aa9ec76"} +{"id": "train_43784", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55faf80d9c551c40e848e3162efc7bb68bc11637"} +{"id": "train_43785", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5f950bf499cc77e3efb651ea185ffadf6429eeb"} +{"id": "train_43786", "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": ["reproducibility", "tooling", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15006074fb4a3803a33c450c7deec2626e4e7cea"} +{"id": "train_43787", "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: 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", "documentation", "auditability", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f7ac0f5d511d33b5f32c4760406568dac3b5614"} +{"id": "train_43788", "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": ["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": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2a20e0af04540166589c112a0b4dd7946af48b3"} +{"id": "train_43789", "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": ["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", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e4da2e42b9e9acc994689608107f02e0c77079c"} +{"id": "train_43790", "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: 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", "tests_are_truth", "tooling", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0fc50c1e14e799965a0df09740b765f1f001e408"} +{"id": "train_43791", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b090e561f9b6ff8a7ab9c2a35e335d7e211d2350"} +{"id": "train_43792", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["auditability", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4ef79b5b420626c37f8a020915993a9d3af648c"} +{"id": "train_43793", "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": ["repo_scale_reasoning", "documentation", "reproducibility", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09b2c2f8aa60b0021373f2fbd0482e5d999117c2"} +{"id": "train_43794", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["documentation", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33c4f6eb3bfaa3a9ef50dee9ed6a8f9d876554e3"} +{"id": "train_43795", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26aaa2a3591160350c12c3bbdc64d2128549b5b5"} +{"id": "train_43796", "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": ["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", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d90244907183d876c52f3b30ca7e5cdf628d3dff"} +{"id": "train_43797", "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": ["repo_scale_reasoning", "documentation", "auditability", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c6856c5871bc8b9f46df5cf33fa1c26a142fcc0"} +{"id": "train_43798", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "auditability", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5136f4f035e7c487a4433fb55c0a8f969fa82ae1"} +{"id": "train_43799", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db7fd564199f8c9cebd3ed92ea9c00b16a8dba25"} +{"id": "train_43800", "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: 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", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d0bcc2f9a31d2794e8879369809f25bd0accbd8"} +{"id": "train_43801", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "documentation", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "685021de37aa3f6b13dabb59674db273fdbcc8ea"} +{"id": "train_43802", "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": ["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", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d01672dd74c62b90ebd1fff46fe25059f615c72"} +{"id": "train_43803", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f5326c774ff3f2c6f6f8e0c0f3ddb0b206f6400"} +{"id": "train_43804", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tests_are_truth", "tooling", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdca544dc1839890e74d2c87379a06b853a18caa"} +{"id": "train_43805", "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": ["repo_scale_reasoning", "reproducibility", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ba735827a3359d0ef083253102a834b0efaa0e47"} +{"id": "train_43806", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56229da2f426577bdbab396ebdf05207b484b76b"} +{"id": "train_43807", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["ci_integration", "governance", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2abaff69fc3f38cbb02470fb20a44710a748b2cf"} +{"id": "train_43808", "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: 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": ["documentation", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80c93df682ce284933ad9eff2b6cd4d94ee9e0a1"} +{"id": "train_43809", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc7d2ca2f5b1b2e456eed1fff7521a430d6f2595"} +{"id": "train_43810", "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: 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", "tooling", "documentation", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d675c2cba3b597c9b9880d59acede0ab783a469"} +{"id": "train_43811", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "governance", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d573f96a24f381f4995779e410c76dd715a869c"} +{"id": "train_43812", "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: 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": ["governance", "tests_are_truth", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c69d76e2ff423969623f822c2b77d30ea82905ef"} +{"id": "train_43813", "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: 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": ["evaluation_metrics", "tooling", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9dd35f83fce02248975208414f371f388243030"} +{"id": "train_43814", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "governance", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0386fcc0d40ff1bca08997a35fc5093ad3d4bea9"} +{"id": "train_43815", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["cost_latency_tradeoffs", "documentation", "auditability", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ef6064becc4967c0a4fa48c848d298afcfa5123"} +{"id": "train_43816", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "415e4c9ae19e12a7a94ec631fd2a63de2fa4aaf7"} +{"id": "train_43817", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9520300739fc841022f8c585eb95e24997d179e"} +{"id": "train_43818", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "943dbc514f5a6e2b7916e8aec4e27abe356fcee4"} +{"id": "train_43819", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3300b958cc3dae8f6001cfc9a14262e0e999862"} +{"id": "train_43820", "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: 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": ["reproducibility", "documentation", "ci_integration", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efb37e37f34f68c277b6c42e4d0e2ab9a65ff216"} +{"id": "train_43821", "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: 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", "documentation", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c6699840bc24997aed7ba9024213d4a2945eb8d"} +{"id": "train_43822", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d40aa165df614e9927eac3738c30c7036d26b27"} +{"id": "train_43823", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20e8873ef7af95b4abdb05f174e0ee29c21f2a2c"} +{"id": "train_43824", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "50667ceb7a120434db2b27b7cd65bab775568b8d"} +{"id": "train_43825", "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: 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", "security_gates", "tooling", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e27738da9c03dbc2a4eff692ffe95ebce08b1aa"} +{"id": "train_43826", "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: 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": ["security_gates", "ci_integration", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8de0e11e91e4483ab3b1e2c5c92c30eb8d63e07b"} +{"id": "train_43827", "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: 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", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e7474d8af74b72bf09b13272240af4445af5fb33"} +{"id": "train_43828", "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: 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": ["governance", "auditability", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4f041ae6e9fb0d60f54449afff186b59b1a3afe"} +{"id": "train_43829", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "895b1e1cc947347695d8e694ba91181d79630187"} +{"id": "train_43830", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d74f862d9bbd07224df128c07cdd4e0c17253c9"} +{"id": "train_43831", "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: 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": ["security_gates", "evaluation_metrics", "cost_latency_tradeoffs", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8212290f9ff89abd4342ce07cd00d7e1938440ec"} +{"id": "train_43832", "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: 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": ["cost_latency_tradeoffs", "evaluation_metrics", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c803bc0b7da890046906ae6530784338d359311a"} +{"id": "train_43833", "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: 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", "reproducibility", "governance", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "042bacd56eafda5c6d56b38f1375474b76dd3259"} +{"id": "train_43834", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "tooling", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cde2d02bc5ebb09fd8c8206bef1287619b2060a1"} +{"id": "train_43835", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83651bd644a0d4faeb46f9741b8e0c9d53c274cb"} +{"id": "train_43836", "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": ["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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a4caea74258787176aa9f6373d993e94b5fc329"} +{"id": "train_43837", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2513b99610cfeff1f724127b4d26b72f527f04d3"} +{"id": "train_43838", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bb3ce4ab60df9a8b78a27b66674a48ef29dee73"} +{"id": "train_43839", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c24f0907b691d631eb9ef3b9e2618eeb9bc5b7e6"} +{"id": "train_43840", "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: 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", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94fbdcaf0bdd668cb300f8cd615d298debdd4a22"} +{"id": "train_43841", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f7e75ff58c5324d75b3962a40bcc1060218aca41"} +{"id": "train_43842", "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: 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": ["auditability", "cost_latency_tradeoffs", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59fe9c3e7f2c9199aa61963c34b286138160862b"} +{"id": "train_43843", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6563466b298ee642558e0752dcac4cc10198c2d9"} +{"id": "train_43844", "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": ["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", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afee596d471c108ecb0c671d64f8c8ea1154daa5"} +{"id": "train_43845", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "ci_integration", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19cb97db7bb8a57bee9f9bd80e266ba0971c5ec4"} +{"id": "train_43846", "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": ["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", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b532afd9cc0f6e9ccccd461048ff45b80882690c"} +{"id": "train_43847", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bcdf849e25afbb9e70c1c0c3a7e8afa4f23e5f5"} +{"id": "train_43848", "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: 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", "tooling", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e89969d7a210786e70ba7f87c91392ae0d6a4440"} +{"id": "train_43849", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37ebb0ac38919803b3dfd759aa0da9e5bb9ae7e0"} +{"id": "train_43850", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8074845951c63e934136c979d8bc816c8b04dda2"} +{"id": "train_43851", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd05373b10fdbc317be5b5b2420c8a362b36cfe7"} +{"id": "train_43852", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f86f0b4b0d09cc2de0c76f6dc25f493711f57313"} +{"id": "train_43853", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a730e119b00349bddfa362c9bb5f77e55f8b857d"} +{"id": "train_43854", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5caa06f55ecaa83a342db2dac23ede7135301d6e"} +{"id": "train_43855", "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": ["auditability", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6d3225c49b9b847f366bb2022f9bfc3a9c67b51b"} +{"id": "train_43856", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["documentation", "repo_scale_reasoning", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c558a21e2e6fb372bb5ba00502ee94dc3da35b61"} +{"id": "train_43857", "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: 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", "ci_integration", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "305ab2356b5ecfaf2a798c8107c0dd075366bf08"} +{"id": "train_43858", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2f6e387fa6453de623d37bfde2cdc2167badf46"} +{"id": "train_43859", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e07c746d42ccb166cc63a72ce0929fc4d189f076"} +{"id": "train_43860", "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: 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", "repo_scale_reasoning", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36c266e6c58409f6661fcfe970c1f82bb244f2c2"} +{"id": "train_43861", "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: 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": ["reproducibility", "ci_integration", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "775c528dcec5838ce47db7a33ba8c70ec5efacf0"} +{"id": "train_43862", "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: 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", "governance", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "849852e9627652d7365bd90ce4ac1d8cd17c485b"} +{"id": "train_43863", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebdf2da374b96f062706d6b58f048b0468ec5a95"} +{"id": "train_43864", "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": ["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", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "917a3c6c12f967dcd1848b24055095a16e5ad604"} +{"id": "train_43865", "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: 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", "cost_latency_tradeoffs", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a63c9dcd20489750c65e552c58d54cad908dc436"} +{"id": "train_43866", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45221b7a2535be00538260f283f1d66bac610412"} +{"id": "train_43867", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ccdc8bb152a2b331e36fbfa663d9b1e02845d27"} +{"id": "train_43868", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31b60596c25c3d75da4c66b8f5605388bdfcf4ab"} +{"id": "train_43869", "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: 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", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e609480a1c7f2ac2fb74d6deef21a664a5b05610"} +{"id": "train_43870", "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": ["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", "evaluation_metrics", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80731ab7f18d7dde875b09820bc850aec3348212"} +{"id": "train_43871", "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": ["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", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb2a9301b8ac14b50ea16fbf2a0837c6f65998b9"} +{"id": "train_43872", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["reproducibility", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c1f0769ce3b4b15d1db1a939295bc63219e7b86"} +{"id": "train_43873", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "security_gates", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec0df10f85a17dd97c4f94f81ea232a04cc2b1f2"} +{"id": "train_43874", "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: 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", "auditability", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3973f4af910fafa9528548e140dc2e6a3d24915"} +{"id": "train_43875", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f24e9fac3822a5995730dbc4107533f83ab1dc3"} +{"id": "train_43876", "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: 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", "repo_scale_reasoning", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47470b708c1f169a6f55a5b91cd3beac7f6488c0"} +{"id": "train_43877", "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: 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": ["security_gates", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84744ec0981db3f23e11769e0469cf195f4c04f7"} +{"id": "train_43878", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "security_gates", "documentation", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5a45c17b352345922cdf7a39e78315c527a7c84"} +{"id": "train_43879", "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": ["auditability", "documentation", "governance", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cb13ce8d15420f9eb018b0f4a5c70ade7a8b07d"} +{"id": "train_43880", "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": ["auditability", "tooling", "governance", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b62ebdd342918a9d72f7e83b953a7bbd2e1ef22e"} +{"id": "train_43881", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc6cd40dc5c7a00c26141b05c405952b6242e534"} +{"id": "train_43882", "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": ["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", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6f897d48f7780574545bef0eba759604564b9386"} +{"id": "train_43883", "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: 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": ["cost_latency_tradeoffs", "tooling", "tests_are_truth", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "128f26c120198428d4c497120a20e0e9bea5710c"} +{"id": "train_43884", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "075e3e09737b09e6fd4d076111d954d29938f5d4"} +{"id": "train_43885", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f284309f1e76d2ce0694c614fcb64705b9b62c4"} +{"id": "train_43886", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22ddb0883242706233a4e9364738ad8376f5f1b2"} +{"id": "train_43887", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "evaluation_metrics", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1330e0ac71028d4d99e8b72ca4f2518d12394158"} +{"id": "train_43888", "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: 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", "repo_scale_reasoning", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42057a934e945af10faf0d22d83376241455e967"} +{"id": "train_43889", "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": ["governance", "repo_scale_reasoning", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe9523f8281f043fe9c56172ab706ef3850ed93a"} +{"id": "train_43890", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["repo_scale_reasoning", "ci_integration", "reproducibility", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9de8ecc0d1bf2788663a05cd3245d6feddfd20bc"} +{"id": "train_43891", "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: 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": ["repo_scale_reasoning", "documentation", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "24f17b79d1ba0d0518c0d90e13a92c76eaeec55c"} +{"id": "train_43892", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01c2a54f043f06b40f89bd010dc8cb9ffb3abd61"} +{"id": "train_43893", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b8be23739189817ad383745cd1a2461e4f98f1f"} +{"id": "train_43894", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1d2b0f7fe45714f87e52909b043f0e70c0ed20d"} +{"id": "train_43895", "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": ["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", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5391df199813e14c5cda6010bf51ecff677e2dfb"} +{"id": "train_43896", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60ba48fa951eaafc76e7d7ef6fe26f0c16738e86"} +{"id": "train_43897", "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": ["reproducibility", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe8e7eacaa608d72cedf07f7fb2f9db230a25505"} +{"id": "train_43898", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "075f443bb5eec0ab9d55ddae81b7a4408687d821"} +{"id": "train_43899", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "governance", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab21195765124b88c88de0a966ab6aa2071ef63d"} +{"id": "train_43900", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "777389df942aed85614fcf46e4f9a524ea9ff6be"} +{"id": "train_43901", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "741b67a9cf16f70bfd54db1626626cd92f6b3787"} +{"id": "train_43902", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a3900c523ed4663420092dd911d3b0201db7a63"} +{"id": "train_43903", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "documentation", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c14e90d0b3ccc8707ef4f849000460575dec5526"} +{"id": "train_43904", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "358969f0207707f764e3ae6691ac2d597216aece"} +{"id": "train_43905", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "708fd225971279ecd5c92850b63ef124b95baf6c"} +{"id": "train_43906", "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: 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", "tests_are_truth", "documentation", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "646530818f0be72511b373b0ad46a831105120f0"} +{"id": "train_43907", "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": ["reproducibility", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46955bb912c535f2d4b75fe67e75c492aed62aea"} +{"id": "train_43908", "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: 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", "tooling", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "782445dedee2374c9414e73077ded41433e38a17"} +{"id": "train_43909", "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: 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": ["evaluation_metrics", "repo_scale_reasoning", "reproducibility", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "781915ad891b6787975174b92c6eaf0de05ed60f"} +{"id": "train_43910", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04d1a638742821a8408746a1fc727c33de1afda9"} +{"id": "train_43911", "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", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6af45377eadaac268716343f3db6393b8e31a338"} +{"id": "train_43912", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "reproducibility", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78f6d0adb083116c61802c9c3cd5f7da0074d802"} +{"id": "train_43913", "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: 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", "evaluation_metrics", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "450ff3e9c51be3e91020ade42baa233e78c84641"} +{"id": "train_43914", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d8482ff0460f081e5222389478714e656e7fbb7"} +{"id": "train_43915", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a350356fee6a2f3e3e19ed44c28a4cbc7684a296"} +{"id": "train_43916", "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: 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": ["auditability", "evaluation_metrics", "repo_scale_reasoning", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2851264525351f589538d1ab6ad1d843f12b06b4"} +{"id": "train_43917", "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", "governance", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67c9df4f08cac10ef6ac24f0afcca7934580184b"} +{"id": "train_43918", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f4af95d2fb4c5cdeb39b563194c80d3328f85cd"} +{"id": "train_43919", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics", "tooling", "security_gates"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18b3244d490a50be156b9f68a71e2c07f71d6139"} +{"id": "train_43920", "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": ["ci_integration", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "514aabf2bab4880ad12c3bc7c9bfdcb7b0137682"} +{"id": "train_43921", "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: 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": ["reproducibility", "cost_latency_tradeoffs", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37bff01c9311cbe3788fdf08bdde43f77f79c366"} +{"id": "train_43922", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "governance", "evaluation_metrics"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "462f90f1ee8d2089ba21840011d1afbec2b23d7d"} +{"id": "train_43923", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06d6d8877bcf59c68741579fc2a5e9135cb014dc"} +{"id": "train_43924", "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: 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", "auditability", "documentation", "reproducibility"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a3e7b8c4df990beb93cf5658d2588094374a745"} +{"id": "train_43925", "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: 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": ["governance", "evaluation_metrics", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3205bffe46d8c53b1824b45368f89b0025532a1"} +{"id": "train_43926", "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": ["tooling", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "644ef6815088d6e866575c4836d212f097e5c304"} +{"id": "train_43927", "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": ["governance", "auditability", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6945b99adaeefaff7a0ee96224c7f38853bae2ba"} +{"id": "train_43928", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38de7fc3aad5f3dc049a740dfc9a6fc318f3047d"} +{"id": "train_43929", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62a583cd38b9ec94dd8c30f52c386a27578d2d2f"} +{"id": "train_43930", "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": ["evaluation_metrics", "auditability", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc23fda4c2757e055b488fa7010c63fd2b72189c"} +{"id": "train_43931", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d26e3d4cd4d201e949d9396e476c26f611293264"} +{"id": "train_43932", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["documentation", "evaluation_metrics", "governance", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc18116889b69656cd9abfe4f2acca8f6d454601"} +{"id": "train_43933", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc85d8b1b9c9b7f872cf250a223bfe5fa03c87b5"} +{"id": "train_43934", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "tests_are_truth", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a64fcfc7a97c1a66ce18c29942149c3f29988069"} +{"id": "train_43935", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "tooling", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "574f2e26aac6453d27f4aa7364056551908b14d0"} +{"id": "train_43936", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["reproducibility", "documentation", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61d9a6a7f69eaf30ba6b1afd5145bfea6f1d4e8f"} +{"id": "train_43937", "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": ["repo_scale_reasoning", "governance", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c00157a1bd597860754197e31366c3f8e4243c50"} +{"id": "train_43938", "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: 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", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4817a5c700c735c2a833d16fff9109cfbfd5ef01"} +{"id": "train_43939", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9b62c1a59096b471a0894486df92eda8394aa51"} +{"id": "train_43940", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17dbd1e5d118c0dda985909f156919afbfdc0f5d"} +{"id": "train_43941", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09c909627b01bb71feb04d3312acc8216b2221c6"} +{"id": "train_43942", "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": ["evaluation_metrics", "auditability", "governance", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68c0e895e704c5915e5e7c6aed39f26df7c879e4"} +{"id": "train_43943", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c37ffbc73de0418a4dc537ba55f9c3a102baecb"} +{"id": "train_43944", "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", "reproducibility", "tooling", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb506d9fed3cfda6180e6439700552033a9e29b6"} +{"id": "train_43945", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f0e3590969e48b89cf6e3f18c897d91a490ce58f"} +{"id": "train_43946", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1ffbcfccd994f02dcddb55ea478fa91bf310587"} +{"id": "train_43947", "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": ["tests_are_truth", "documentation", "governance", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a14fc6373cb881c2dd071e1dfc773dff85cad634"} +{"id": "train_43948", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["governance", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e1f25f8216675a76e0a36de7e4e0b2688c92baf"} +{"id": "train_43949", "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": ["tests_are_truth", "repo_scale_reasoning", "tooling", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "03faa0bc234fd8de50fcdde9518c4538826b5d72"} +{"id": "train_43950", "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", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d5c5626db9b5971c70e5949244e9763ebf4fa2a"} +{"id": "train_43951", "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": ["reproducibility", "governance", "security_gates", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4bf4127b3fbdd58db970dbcd2c191737212f713"} +{"id": "train_43952", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "auditability", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25324826198ffa746f0567e85e8e7f84be076d5d"} +{"id": "train_43953", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2512fca105cf411028a33988ee3c1184ae478d51"} +{"id": "train_43954", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "governance", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a48d85509cc98e6c0247c62339e59e73c05f87bd"} +{"id": "train_43955", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af7812f82e2a96e1cd8f2bedd69d72c3d39ed6c6"} +{"id": "train_43956", "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": ["reproducibility", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5452a28d34c917b7e1f47604f897a8e16ff5b114"} +{"id": "train_43957", "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: 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": ["documentation", "auditability", "governance", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "704bb005731101e7369696ba87f47553af503f7e"} +{"id": "train_43958", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f396a87f7b186f37e17e0704c8ca43c330be1a2f"} +{"id": "train_43959", "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": ["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", "auditability", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c689137c0dcd8b5540af1550fa00a46b5f4adf2"} +{"id": "train_43960", "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": ["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", "tooling", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "618f97f315e3fc62601c574a44e3111012f37b3a"} +{"id": "train_43961", "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": ["security_gates", "cost_latency_tradeoffs", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ab18a19999d208c7340030022efa2651c850b31"} +{"id": "train_43962", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98dfce3a6aec6a60fa9f54d8e900c1f60f40b618"} +{"id": "train_43963", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cec4d00eeb0dde799495cde24e6f711adbd9b2f9"} +{"id": "train_43964", "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: 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", "documentation", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79c710cf6c44f068d11025e43dab312be40cbe05"} +{"id": "train_43965", "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: 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": ["auditability", "documentation", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c824859951bc8b25d20c3c8d904b0c9f16bd0876"} +{"id": "train_43966", "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: 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": ["ci_integration", "cost_latency_tradeoffs", "tooling", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "78111ccd211f16c6002b1547cade30c5a97ce9dd"} +{"id": "train_43967", "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": ["ci_integration", "documentation", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1a6d43be1e78ceff0d95bb4f92dec2eb92138a8"} +{"id": "train_43968", "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": ["repo_scale_reasoning", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8ec6c73e90995a0922ec1fae6e6cfe6349ecdae"} +{"id": "train_43969", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29cb455cb46eee6e0a8174bdfa59e2c3e5bce77c"} +{"id": "train_43970", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a76d8f024f23b51036ee3d41027e0d3c5d411364"} +{"id": "train_43971", "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": ["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", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a40cb459224ae5640510babf1701ab8fda9c7841"} +{"id": "train_43972", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "102737d0d26a3a6ff789059a31299546a73f4cf0"} +{"id": "train_43973", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "tooling", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbe1061dd28b08412a7f2bfd3ebda7c60da577ca"} +{"id": "train_43974", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7857005340fb94de39f5ad336cf1cfe82d26ea37"} +{"id": "train_43975", "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": ["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", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7379cc947adacf57c0fdabf1c31945d031dfff35"} +{"id": "train_43976", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1cd2ca3b30521db7bb1bef09e1f105bf2f691571"} +{"id": "train_43977", "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: 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", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "006387ceb4bc165c398f216dc92dc6fdbfe3030d"} +{"id": "train_43978", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9af7f6b1dafcd90ee034d9272a915f33ffe2e885"} +{"id": "train_43979", "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: 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": ["auditability", "security_gates", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8433b93806187746fe182fc2cf2b29977e0ddd31"} +{"id": "train_43980", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "reproducibility", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af4fa1a2d6c7a396907c41069674ef17fe9a738c"} +{"id": "train_43981", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "evaluation_metrics", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "381b9b4705652c197731c777aeb7cd015af4b0b1"} +{"id": "train_43982", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "209c4cbfc75314b575a9dab31556f074de6bc51f"} +{"id": "train_43983", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd6951e2d6da2fd4a1bd2862df075ccc5f768d67"} +{"id": "train_43984", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "278f08c98d524a3d59f2d26e8ba0b4088afeb41b"} +{"id": "train_43985", "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: 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": ["tooling", "evaluation_metrics", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e4ed8e285994a032f39f3f3c303d75643aa9f31"} +{"id": "train_43986", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "governance", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8bd3101e14ffe08e93906a4c067d6f22a6632a4b"} +{"id": "train_43987", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "385d309ddd322c74a7802a28329bbad05c43a855"} +{"id": "train_43988", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "auditability", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc8bf40081ea6a5b2726fbb232987788d50a6446"} +{"id": "train_43989", "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", "governance", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1a389580ce94179a394494f620aeba702103549"} +{"id": "train_43990", "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: 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": ["security_gates", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3d1a6a29d597e9567ade3c72d357854da9cc81c"} +{"id": "train_43991", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3ba53d1786a4904c3cfd9b889d2058fff712ee4"} +{"id": "train_43992", "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: 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", "auditability", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "110c0f8ea05388e34b970b6ff0073a344615076e"} +{"id": "train_43993", "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: 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", "auditability", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf04ff62fec1ee2fb55e45feccafaf4999513fee"} +{"id": "train_43994", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["tests_are_truth", "security_gates", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f66903cbb736fb5776799dbb1c75cd651d57dddd"} +{"id": "train_43995", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a0971055b5beea88b62270ac6c18ba387884aeb"} +{"id": "train_43996", "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": ["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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad5673c2a037a8fd3b012788a3d668ecc4f5557f"} +{"id": "train_43997", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf93c712c7b9931c51ee2a8b9634eee5c8f4cbec"} +{"id": "train_43998", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["tests_are_truth", "governance", "auditability", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3665d73eec4caa5f6bc548223a5d025346e396d3"} +{"id": "train_43999", "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: 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": ["tests_are_truth", "documentation", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "268e41bc8fbb5397824dc23c175066f15cc2b98c"} +{"id": "train_44000", "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: 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41c64226b67773a3f580ee296ecbc141688127cc"} +{"id": "train_44001", "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: 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", "reproducibility", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "323e6993c6fd01b53c93d3338c6209dbfc60e96b"} +{"id": "train_44002", "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": ["ci_integration", "evaluation_metrics", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a8d43c705c8db205923c12b7e83422eb335309a"} +{"id": "train_44003", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d22cc9ad34f35110dee944567bc0037b4c391731"} +{"id": "train_44004", "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": ["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": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb1ddd399277687f9948d266fafdf9e2ae3a74bb"} +{"id": "train_44005", "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": ["security_gates", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9768452fbb74886295779a4a8f0ba320fd2aeea"} +{"id": "train_44006", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "governance", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25458ef71424d9a79fa2c1f1eac2519babba3623"} +{"id": "train_44007", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7081bd64e6714bd2940b6550a70ed1249c268b1c"} +{"id": "train_44008", "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": ["auditability", "security_gates", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9dbaf3ee9572b3f2e9d33453c0e7df3a3923af1"} +{"id": "train_44009", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d679c205be01274bc44238e04e8fc9a327c53c6b"} +{"id": "train_44010", "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": ["ci_integration", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "893a5fc22bf0fbd67e23bd01cc32733005cbb8d6"} +{"id": "train_44011", "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: 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": ["auditability", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d9fc10b97d8e3c04e043c9e08b8de6ad4688b5e"} +{"id": "train_44012", "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: 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": ["ci_integration", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e206dcff951e433416ef57887530471ea2eefbb"} +{"id": "train_44013", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f494462882bd7a915f590baf720ddc60bc9eae9"} +{"id": "train_44014", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ab1242e9570313804a4f5fa69d13592ab22bfbe"} +{"id": "train_44015", "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: 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": ["evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09dd1c267b297ba01ba15bf4afa30e6732e93f48"} +{"id": "train_44016", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6df9328be8294e3df6f1734fe57f814225445865"} +{"id": "train_44017", "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: 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", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ebc897824efa75147612af6472b0b63e0bfefdb"} +{"id": "train_44018", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "reproducibility", "evaluation_metrics", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fe57b49dea57692efca2ef7f110d2b70b9b4f57"} +{"id": "train_44019", "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": ["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", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5861c44ae0af941b68f3e713b5ea7576a7a60210"} +{"id": "train_44020", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18a57ee71853d678cd6c31c7f49b2f0348b74e7a"} +{"id": "train_44021", "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: 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", "governance", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d12d6fd6d204a1bb98ac1a89e6e8669edc6b33c"} +{"id": "train_44022", "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: 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": ["evaluation_metrics", "reproducibility", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ebbf6e82860f953474251bf0fed4ba3e1be7087"} +{"id": "train_44023", "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": ["evaluation_metrics", "governance", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "514aabf2bab4880ad12c3bc7c9bfdcb7b0137682"} +{"id": "train_44024", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "documentation", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0aa4c6a68ab37fa8977cfd22a283ba61809fdde6"} +{"id": "train_44025", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "ci_integration", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14978c48dea43d67202945c79b4888fad5187355"} +{"id": "train_44026", "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: 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", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "820e4362c5d9055d23e33a90f1a10b48dc13a7f0"} +{"id": "train_44027", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "reproducibility", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7639325a30eb01fb385c2554c37305821a2be074"} +{"id": "train_44028", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "799ccd85c328575f5786d8566c41f56eb73471bf"} +{"id": "train_44029", "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: 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": ["ci_integration", "governance", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6577ac2b2fb6e979befc579b5b9b7159b7f3da61"} +{"id": "train_44030", "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: 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", "auditability", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c815aa1d9114ea91fb4629bfe58d6f2f7f2d7b8f"} +{"id": "train_44031", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c466d7414a57f8c1c20a1809dea1477d71153cd"} +{"id": "train_44032", "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: 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", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52ac5432b0e43fb524d3fa95ec86d9cdd4e4090e"} +{"id": "train_44033", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "053ffc2cd05af22c7391d8bf3073910ec3fa54ec"} +{"id": "train_44034", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b480272d9013249b4cdb618f6481f08bb5bb0ad"} +{"id": "train_44035", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d467e0e9226df253aa4c1083a3f2f3e6dc333e40"} +{"id": "train_44036", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bbbcb56e9146d9307e994fd6af9e038118e5485c"} +{"id": "train_44037", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eab970f17063ea791dfd9529a7c1c1d9f5efd8ba"} +{"id": "train_44038", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94eece7310d16192f4b552e98556c8198b83682d"} +{"id": "train_44039", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "governance", "reproducibility"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa0b47c43f0da7aea307e4f68ae3431e18b277f2"} +{"id": "train_44040", "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: 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", "tests_are_truth", "reproducibility", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccc6b7592c47b722035bf0ccde2d0bf1e45ca5f5"} +{"id": "train_44041", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e64b37d4c81548c502096e515c72c23e93ff45cb"} +{"id": "train_44042", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "df630987ac2a0881482bc88a618b37190daeaae0"} +{"id": "train_44043", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0873ef5dd0407ac456f54a096461d1a879a8e4c1"} +{"id": "train_44044", "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", "documentation", "tooling", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b0170cb245c2e86099f7a3757bc141f4ac3a8ea4"} +{"id": "train_44045", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "documentation", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28bf92026ad43679236ea301f97c41dd27a9cd0c"} +{"id": "train_44046", "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": ["cost_latency_tradeoffs", "documentation", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f309d769211b31ce082fbe9b7bc12e3c7f34da7b"} +{"id": "train_44047", "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": ["documentation", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecf43aab777ca4347221fae797bd13d58fac9dac"} +{"id": "train_44048", "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": ["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", "auditability", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d90244907183d876c52f3b30ca7e5cdf628d3dff"} +{"id": "train_44049", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "tooling", "governance", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb8a37fd91d744a7bb38f7cf834855d58fed552d"} +{"id": "train_44050", "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: 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": ["governance", "documentation", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "90f0c30e4562d8c169841d2c9f256aea23ddbc17"} +{"id": "train_44051", "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": ["governance", "security_gates", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3682cb717e97be48e7853a32f4d93920e62015aa"} +{"id": "train_44052", "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", "repo_scale_reasoning", "tooling", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f82b4a683ebbf15b09cf8d20104c4b7d55c9646"} +{"id": "train_44053", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "documentation", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3816932c8810379f948ffafd45527e1f18db769e"} +{"id": "train_44054", "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: 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": ["security_gates", "reproducibility", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52710a70e929185d0ac959009a29a67d577d0549"} +{"id": "train_44055", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "evaluation_metrics", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "441e2f9e13fd14c77efd7955537aa3b731787081"} +{"id": "train_44056", "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": ["documentation", "auditability", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "492c12e8a64e2de8ea4c56ced652485430c01f9f"} +{"id": "train_44057", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d80e6c568dc2203426b04de48f1b33ef15816842"} +{"id": "train_44058", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "governance", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f12aad9b09169697646aa84a1382caa8381ae025"} +{"id": "train_44059", "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: 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": ["evaluation_metrics", "tooling", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63893d19fa45a7e30a3ddbbfcc771325fc8ad030"} +{"id": "train_44060", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["governance", "tooling", "ci_integration", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "938652c27ed3074b2fcd514ba569152a5ef5b169"} +{"id": "train_44061", "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": ["tests_are_truth", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "431bc81728b4f5f96f8b7fd1ac4a855cbf79d3c2"} +{"id": "train_44062", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "457261f137a9f5fca88314ff482cc400ac7a5aea"} +{"id": "train_44063", "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": ["cost_latency_tradeoffs", "tooling", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57c38ec0d62d59d82d625c84f9355b7de20f46d8"} +{"id": "train_44064", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc735220db6fe5787d570ddd1d27136572876bf9"} +{"id": "train_44065", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87a2eca5e9d3f44f1216d6fa2189ff8d53a5f76b"} +{"id": "train_44066", "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": ["evaluation_metrics", "ci_integration", "documentation", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "461df607e26b8404a531eb9a67314cd1ab6b2b76"} +{"id": "train_44067", "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: 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", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "63596d09071f39033c7416f284da75d8451b85d5"} +{"id": "train_44068", "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: 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", "documentation", "ci_integration", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aac2ac08171e5392c7f006dca3cffc87e92512df"} +{"id": "train_44069", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cfc096b8c10b54853acf13c70b341eae118ae43"} +{"id": "train_44070", "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": ["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", "documentation"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce1c12f977f3f3420e858356ccbebccc95d59f49"} +{"id": "train_44071", "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": ["evaluation_metrics", "cost_latency_tradeoffs", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58db8de21105f42b6f7a69ba7e2253d275683f38"} +{"id": "train_44072", "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", "cost_latency_tradeoffs", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7606cf7bb103a4abad41ad28d6b6ec1d6b58a3cd"} +{"id": "train_44073", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f9af930da225d259f25330331f08ad0ef76765a"} +{"id": "train_44074", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "security_gates", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13fb2688ab509d8e3a67177d939139e4378a2787"} +{"id": "train_44075", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c6fb2e5209da2d5a55d8133fb1a2892baa16b16"} +{"id": "train_44076", "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: 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", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d93c9a556a8809c750426d97d9c05c49730a209"} +{"id": "train_44077", "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: 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", "ci_integration", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80e749d1221eb9171002559625f5da33c63f5e6d"} +{"id": "train_44078", "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: 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": ["ci_integration", "auditability", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8221472930450fb168f4136bd89732645971611e"} +{"id": "train_44079", "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: 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": ["reproducibility", "security_gates", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e17714401e583a8b0ca31a0816d5a536acff615d"} +{"id": "train_44080", "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: 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", "tooling", "tests_are_truth", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "385a24df1d6093cd4c2b0ff75f60b5ee12b578e0"} +{"id": "train_44081", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d35c085e93f49917ab196296906ee21dc5b6ffe3"} +{"id": "train_44082", "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: 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": ["evaluation_metrics", "ci_integration", "documentation", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "279227a82e65bdd4549abc601f6a029b126ab61b"} +{"id": "train_44083", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "documentation", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6462b9156048ef7dca59d73ff39f5cae90b580bf"} +{"id": "train_44084", "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: 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", "documentation", "ci_integration", "security_gates"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d953c8acc87ad274d04372b3eed490a631030c1e"} +{"id": "train_44085", "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: 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", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a18c3dce90a88bc75aff8eddd3ad45b1d0e6826"} +{"id": "train_44086", "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": ["reproducibility", "auditability", "repo_scale_reasoning", "tooling"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "559ccda68ba4586d79396f350cfd81a350f7200a"} +{"id": "train_44087", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "tooling", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e00b3c208bbeb2af75ee8b6861b79e7e1984889"} +{"id": "train_44088", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6936664fa8483934989ae61f9efe413a188d138"} +{"id": "train_44089", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "documentation", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f733e1f98bd72e5f5bebb1795dde93fbd1fbafd6"} +{"id": "train_44090", "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": ["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", "auditability", "repo_scale_reasoning"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c2c939d655f1fec4fae390fef08a56aff7109ac"} +{"id": "train_44091", "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: 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", "documentation", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d7d19da58aa6fa8df7d5fdabdd2698c88611ca5"} +{"id": "train_44092", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["ci_integration", "reproducibility", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a8b17b1cb74705fb2940a0c8fb16a70a73d2581"} +{"id": "train_44093", "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: 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", "ci_integration", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e8e0fa741bc59173a47f82c1af5dae7b8f1bf2c"} +{"id": "train_44094", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cc582385ff569c0f8e90dca0f97474c48be578f"} +{"id": "train_44095", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb1cf62ce85b77364d632fcef9fe9a3712c674a5"} +{"id": "train_44096", "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: 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": ["auditability", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b54b13d176f21b68cb72c43e522520db10c6a3fd"} +{"id": "train_44097", "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: 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", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "877ce1a5a3dba96fc1ce4b602e5008fbf061bc46"} +{"id": "train_44098", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b47efe4e499a2cc3f76e4deaf49cf0118350f4f6"} +{"id": "train_44099", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db7fd564199f8c9cebd3ed92ea9c00b16a8dba25"} +{"id": "train_44100", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79effe8d8f23e2e4b7662cdf2ae086dc899bcf05"} +{"id": "train_44101", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a6be66bf5152b2a52f1b2f3c83067cb1e6ad9f7"} +{"id": "train_44102", "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: 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", "security_gates", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3775a586e1096557ae8e709af7e5f197ead18d17"} +{"id": "train_44103", "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: 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", "security_gates", "tooling", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "589fa069e9fc150bca03ff820d3f29a05972090a"} +{"id": "train_44104", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "09b705f8b75dd5d411c21394691d79e230692c86"} +{"id": "train_44105", "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: 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": ["tests_are_truth", "documentation", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4607894d9b7a300f55d78df4c8aed7a6b39fb575"} +{"id": "train_44106", "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": ["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", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62bdb40fb861a019bee08c2564234fa5e3dfdd0a"} +{"id": "train_44107", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d5942a4b616b0830c20cedc655bd8b5adad8d93"} +{"id": "train_44108", "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": ["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", "tests_are_truth", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62cb8b9a9bbd185da8f87869018dc95dea179223"} +{"id": "train_44109", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc9932597b0cdfda9a74234231a93d842045b5cf"} +{"id": "train_44110", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0bb1cb3b4d3af417ab7d74585ce04dc1b0f38d19"} +{"id": "train_44111", "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: 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", "repo_scale_reasoning", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ebc897824efa75147612af6472b0b63e0bfefdb"} +{"id": "train_44112", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "security_gates", "auditability", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdeefee183582262f88206a4e347b27d69f2e5d8"} +{"id": "train_44113", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6e75f278064879c5219744dbf586b3e1dd5e6bd3"} +{"id": "train_44114", "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: 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", "ci_integration", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "758412a33208cc61340f085db01f21e30c9cef05"} +{"id": "train_44115", "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: 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", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d504d7474dfa4ca7343e2d1ced550caa9647c4b6"} +{"id": "train_44116", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready 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", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c927cafaa15686c965761eeee2dc03879d6a213"} +{"id": "train_44117", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "reproducibility", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "528175710d65be7374506190593b5ea4e41f6155"} +{"id": "train_44118", "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: 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": ["reproducibility", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3007d829e626b963939e0fe33244555974421461"} +{"id": "train_44119", "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: 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": ["tooling", "governance", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe7c7632d1341aecbd7e75354ff10a871eb85454"} +{"id": "train_44120", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8d382fdfcc3b084546be8ec49a912524b21edc6b"} +{"id": "train_44121", "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: 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", "security_gates", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcd5efd991cc902bc4f1cd5710762fe61b05d191"} +{"id": "train_44122", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b0152fc867cb73ce7f0a370c4069462bcf7cbc2"} +{"id": "train_44123", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57abcb1194ef9fc8014bc9b8c30a203b466de7fd"} +{"id": "train_44124", "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": ["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", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6762e79c9a68ea273180876043559ce258666ae3"} +{"id": "train_44125", "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: 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", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5cd521284c498dce77f6b49f132e6489a1b1604"} +{"id": "train_44126", "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": ["reproducibility", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e8340b7c056f9fb55a075b9ed42859c5e966591"} +{"id": "train_44127", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27ade62b1a4083f062ce61f9254371c3cb96794c"} +{"id": "train_44128", "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": ["repo_scale_reasoning", "tooling", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73d53c9590ee4deeac43dc281a4d088c8effcf82"} +{"id": "train_44129", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21498f064008547e0b14227aa55344e8f72f8e62"} +{"id": "train_44130", "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: 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": ["repo_scale_reasoning", "governance", "documentation", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77db71fddf5bde859f7b4bc4a14e3ead264b07e0"} +{"id": "train_44131", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["security_gates", "cost_latency_tradeoffs", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ce247574c925702f8ddf7b324b0742d4ada6b0d"} +{"id": "train_44132", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d582a85c87f3e200c55fcc2a61ad4ab710dd0b5"} +{"id": "train_44133", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b28e8a8367da16895aa81705470b5fd21ba33f0"} +{"id": "train_44134", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "631eebc58e8484db95cba88b3b4ab2a14a0171f4"} +{"id": "train_44135", "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: 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", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e2ed6cb2f0554868c148743af3783c619729d20"} +{"id": "train_44136", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5ff58f871f6f73dbf6750234a5fce9c6cb0fc85"} +{"id": "train_44137", "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: 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": ["security_gates", "ci_integration", "documentation", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8ed8cf037d33d9d36ab8d540c63bda0200297e1"} +{"id": "train_44138", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bab7cd3b1aacb4de1e9a5e47155856b54291671f"} +{"id": "train_44139", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3441b76da53ef844eefe3cf630a34d3bbd8ca84"} +{"id": "train_44140", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "475c4f7094e1c071bf03f48ab3a16bea4e597414"} +{"id": "train_44141", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d7fa9ab94ad6516fe2af36957b8dcb9bc95f171"} +{"id": "train_44142", "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": ["governance", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "591fb956552f22f601423eadf4872803cd4299f5"} +{"id": "train_44143", "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": ["reproducibility", "governance", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72cde4bb62c0ecfe135382a7ebf4486b79174688"} +{"id": "train_44144", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c17f1bb9bdeb666191f0fd26c269e04da2038b11"} +{"id": "train_44145", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["evaluation_metrics", "ci_integration", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7a3031d97220ffdfdd40e4c7032768a14ac83d1"} +{"id": "train_44146", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff7dab6c092628eaa16c4fc472a3554c171ee420"} +{"id": "train_44147", "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": ["documentation", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2000560a3532c59d9e88e3fd110df792183f5237"} +{"id": "train_44148", "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": ["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", "ci_integration", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2a76788effe340e2217b7e443fc236d88857fc4"} +{"id": "train_44149", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["documentation", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d559bf3f91286992d55d6679105f32afcdbd13b4"} +{"id": "train_44150", "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: 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", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ade0c0dcbdc1ee19b9af0e484890f1c3a2399cf"} +{"id": "train_44151", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49dc16a4f4f9c1d78ea09524b3fb7c312cb78ddc"} +{"id": "train_44152", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7accb4d6051e4ec1a92f8b97956351786ddb174"} +{"id": "train_44153", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["repo_scale_reasoning", "ci_integration", "security_gates", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "511bb8241930aeb214d399bd67edceb5e2330d15"} +{"id": "train_44154", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "repo_scale_reasoning", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "635aeffc6f52fdb48fdb948b5ed9af400f572e17"} +{"id": "train_44155", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6b618306205e6534d285fbd95b71942c39acf80"} +{"id": "train_44156", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5e97295d7aba31b4032027a152d55c8e5bd6f91"} +{"id": "train_44157", "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": ["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", "documentation", "security_gates", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa1db1df4d64f97f3ae6709757f75d686de161b4"} +{"id": "train_44158", "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: 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": ["security_gates", "documentation", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "240a65b05452b964c4566014741d5297f8b2b750"} +{"id": "train_44159", "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: 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": ["ci_integration", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13a20457ca0a09986a23d9be330a9612f2575b86"} +{"id": "train_44160", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f4bc176a47124910070f9d3450be12f0f20d966"} +{"id": "train_44161", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f54dad4831a2058bf6a15183187fc24987b7fab4"} +{"id": "train_44162", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4db337c377d8ca6ee278b18fec98ea1bae97c79d"} +{"id": "train_44163", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9d1841d0c7e2e1b61e956bd5b12844c0d6d5d0c"} +{"id": "train_44164", "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": ["cost_latency_tradeoffs", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f3b82c3b46ce5151983fdf71c7c46aa7b245ed4"} +{"id": "train_44165", "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: 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", "cost_latency_tradeoffs", "governance", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da42d9e0d6df0d76a42a2c768649e99f14acde70"} +{"id": "train_44166", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ed9f3598c4379cbefcf71c698e79591e2ea5b44"} +{"id": "train_44167", "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": ["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", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5891b30153211125ea811bb0460c83c6b62935ee"} +{"id": "train_44168", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5dc29d3ed1172d9d19eb767f53857f71d749b84"} +{"id": "train_44169", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["tests_are_truth", "security_gates", "auditability", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9f19113b3d0cc319b9a4226c9b20f35a8b6e700"} +{"id": "train_44170", "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": ["tests_are_truth", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c445b07747f4052d15bc3fa4fc7ed14ba62aef4"} +{"id": "train_44171", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "cost_latency_tradeoffs", "governance", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fb91e9cc17b75d690bdc40473808ef118719e355"} +{"id": "train_44172", "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: 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": ["repo_scale_reasoning", "tests_are_truth", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38a6fee21d5cc91b251982d8332720eaaf2bc9c5"} +{"id": "train_44173", "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: 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", "evaluation_metrics", "governance", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b9ca11d22128e41dd9292191a499d3b22561ab71"} +{"id": "train_44174", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9b0c6c05e4c44a30ec17ff5a165782f7d2587a29"} +{"id": "train_44175", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "documentation", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cbda5b5948afb982797e8143043b1f34d086d61"} +{"id": "train_44176", "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": [], "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": "c1ed93264bf4fdfe138d0c4f879f0f1a363e246b"} +{"id": "train_44177", "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: 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": ["tooling", "repo_scale_reasoning", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08f24d1126f06b866a35a3c041765c46b1a648dc"} +{"id": "train_44178", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afc494647434a93eae23e35f324087de431457d8"} +{"id": "train_44179", "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: 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", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad9ca42d6a7fdeb41c8156d243ec6d9a4bcf010d"} +{"id": "train_44180", "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: 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", "reproducibility", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4befb2c4376feaaad5418298be9fde7d3eba1ca0"} +{"id": "train_44181", "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": ["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", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38e75bec6fceace823d7d0acf78cf8272d0384de"} +{"id": "train_44182", "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: 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": ["governance", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e2281e720cb845a86b2f4f87ffeaea3304595b3"} +{"id": "train_44183", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "577230705a404d0a3d095e6941524c463c377f1d"} +{"id": "train_44184", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cafe257c6eda478ac5b034f17c4f7c06fbe36b4"} +{"id": "train_44185", "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: 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", "security_gates", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ec793e74b7a77238d196132885fd66098eb84bc1"} +{"id": "train_44186", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["auditability", "governance", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c3770e7a94151089c6377d1dd1956a798f6b87f"} +{"id": "train_44187", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac73655abc1c42d20079237b0de78b45154c1011"} +{"id": "train_44188", "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: 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": ["ci_integration", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "298ada66a996e821fe13f0b85b2a2067b4652753"} +{"id": "train_44189", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2f00e56169d6a9c40c564f6f4672c831ae62aee"} +{"id": "train_44190", "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: 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": ["repo_scale_reasoning", "security_gates", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9d2631b90a5775b4111b9dc73ec72c0727ce640"} +{"id": "train_44191", "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": ["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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2684f44cf4ea522415d5342aadeabbec17f5db6f"} +{"id": "train_44192", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "auditability", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87ad400fb436cce7274384967d6e1714d4b9c339"} +{"id": "train_44193", "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: 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": ["tooling", "auditability", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "077e8062b2b86c4574aa29f3e1412ad415e0cfaa"} +{"id": "train_44194", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c91c66799b2a2aec58dd716bbbe798ae74b8ea9c"} +{"id": "train_44195", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "governance", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a8eb02b74cc419003d73c5e172e67be10de9c38"} +{"id": "train_44196", "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": ["reproducibility", "governance", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "401210cb69046b5c8791915354e889b20ca9eeb4"} +{"id": "train_44197", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8764705f751db6c529c210f10fca2da487e4c93"} +{"id": "train_44198", "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: 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", "auditability", "governance", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4386d42a8ff0f9e567f300ece0a437f6f1fa482c"} +{"id": "train_44199", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bebcd6b8f6f87714dc3a1cab8bcc62bef8fec950"} +{"id": "train_44200", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce455e75fa6aed9ac04eb5d2824370e2f857a836"} +{"id": "train_44201", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c110d36471bfbb79dbfd23490e14b042c4dd063"} +{"id": "train_44202", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d8c1ac200c74ccb6232e3a219788901dd4510a4"} +{"id": "train_44203", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "auditability", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dafd628bb45541500f6d6caeca1fdd7cb9e24e0f"} +{"id": "train_44204", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "security_gates"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a7b7698c48f20d00593097bb7b075fddb0cb8256"} +{"id": "train_44205", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89dba65c39c16c84ba3694c7a1bffaac6f906acd"} +{"id": "train_44206", "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": ["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", "security_gates", "auditability"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5840a595c29379955af6d79fcd45941d7d67fb3"} +{"id": "train_44207", "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": ["documentation", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a28b345e3c712f544ca7d62453c98580642d8f94"} +{"id": "train_44208", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dec44cc5b7f03d6c332ef86bda30481b83ed56e0"} +{"id": "train_44209", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "auditability", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "619f21c0be1270e959f25314587a27d455a9aa52"} +{"id": "train_44210", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31717dde7d7dbcdc0c2d2d6013c4fbfab9159d83"} +{"id": "train_44211", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "137ac4c8989fa60dbc00c3b9858b13dc49aef689"} +{"id": "train_44212", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04ba8f290e4e4ffb76f0724b694e5d01e889aca2"} +{"id": "train_44213", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "governance", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a4648ce14272fe09995680aa6600fa855db38cf"} +{"id": "train_44214", "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: 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": ["ci_integration", "repo_scale_reasoning", "auditability", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3854e28e9494a690586478020b8f1abd67da02c2"} +{"id": "train_44215", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "governance", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97cd3224985c5c2ff2746f0e9c19ffa17b634536"} +{"id": "train_44216", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00b8c37535032a7546c3f78118a34a80a926e8e7"} +{"id": "train_44217", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "tests_are_truth", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8a1c76043cbf3b355d1ac493bd0105ca4679080"} +{"id": "train_44218", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "873c174b40a5cfe5bfc0c930c8dde9a1c6dbb405"} +{"id": "train_44219", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87ab6de2046081238fad9c7c66aa350b7ae2656f"} +{"id": "train_44220", "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": ["repo_scale_reasoning", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12886fc7bcb6afc74fdc47092ee07e4c853dd4ca"} +{"id": "train_44221", "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": ["documentation", "tests_are_truth", "tooling", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ce28d64eea86d6aa01de56454b3da47aada1d3d6"} +{"id": "train_44222", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d36a2ebbeaaaed0adc828cbe56fc9dd41c845b13"} +{"id": "train_44223", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "governance", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a97d7d754135f0639f3429efb09c5124367982a"} +{"id": "train_44224", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tooling", "security_gates", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "293fd1f09857e73612b114fc7f76789899057c11"} +{"id": "train_44225", "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: 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": ["cost_latency_tradeoffs", "governance", "documentation", "auditability"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e4189a2010121403e6e742484f9e53efc2e1ab2"} +{"id": "train_44226", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "governance", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a082729be68cb51fc7ea4e64b782131b850082ab"} +{"id": "train_44227", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fe866f1f66a66dd42b4035fe7b00dd57700d334"} +{"id": "train_44228", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "tooling", "documentation"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28bd68ca16324f00515312881ffa1f036196c17c"} +{"id": "train_44229", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1f2a0950656faad5cb273ef5677ed1a6644e89af"} +{"id": "train_44230", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4147bfbc90267075c1f728a05e8bb6e3e54b9987"} +{"id": "train_44231", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "documentation", "governance"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6a85cccf6ab019a882cccfbd963171fa8e024c8"} +{"id": "train_44232", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "reproducibility", "governance"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27ade62b1a4083f062ce61f9254371c3cb96794c"} +{"id": "train_44233", "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": ["tooling", "auditability", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44b0ddd5101a7c78b6749185e54792ea63c7dea6"} +{"id": "train_44234", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "documentation", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a0971055b5beea88b62270ac6c18ba387884aeb"} +{"id": "train_44235", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef0674acc8a47591a589bc49a236574b710d2af6"} +{"id": "train_44236", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "repo_scale_reasoning", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12870ecd6e63c195c8e3f932c1341db3b3752ae2"} +{"id": "train_44237", "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: 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", "documentation", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e33489389ec85818e8a46b3c2c379528fe967de0"} +{"id": "train_44238", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e87f05423efd3f9b9d7e3982e13df6d492aa90c5"} +{"id": "train_44239", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "tooling", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d679c205be01274bc44238e04e8fc9a327c53c6b"} +{"id": "train_44240", "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: 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ce7d4ab981e154d9b3a32e1e4ee7b6f1498b9bc"} +{"id": "train_44241", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "tests_are_truth", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2233141afe5b15ba1d7228c5207055d7cef58ab6"} +{"id": "train_44242", "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: 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", "tooling", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f3929f2518fb07838db64f882384ca2265bbd4e"} +{"id": "train_44243", "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", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f23ecb5e6357c5f1ea81f216f617dea2c2084701"} +{"id": "train_44244", "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: 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", "auditability", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3699e7968ef42db6906d3ebfab8a47136015a3de"} +{"id": "train_44245", "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: 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", "repo_scale_reasoning", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2398b0e383b00af4545369ae2b435ceb79e17969"} +{"id": "train_44246", "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": ["documentation", "tests_are_truth", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a1d09aee2c4c43a08632cc47291363a67a06ab18"} +{"id": "train_44247", "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: 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": ["tests_are_truth", "security_gates", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd8438307434759d856d203b978ba5084cc5f62a"} +{"id": "train_44248", "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: 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": ["auditability", "tooling", "security_gates", "governance"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4829a867e747d26063c9b17a5a0a47237751ec0"} +{"id": "train_44249", "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: 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": ["security_gates", "repo_scale_reasoning", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4baf668c8f899d3518f53845fcd1fa885a82b6d"} +{"id": "train_44250", "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": ["repo_scale_reasoning", "governance", "auditability", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f99659e763e465e95cd4ebed0b1078bd815e62ac"} +{"id": "train_44251", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "270cdb2aa5c632498915559f8a0ec2f02958c0a8"} +{"id": "train_44252", "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": ["security_gates", "tests_are_truth", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5ed8baa86bbf24cf575eb40a92f1075e94d4a19"} +{"id": "train_44253", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc73c4b27ef6b31b2ec082bfc97635c3a8b349d4"} +{"id": "train_44254", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7952105f89cb200f4bab609a353ec65e688a0dc1"} +{"id": "train_44255", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0b5418be80632f6fb6aa5d619105c95b1e46220c"} +{"id": "train_44256", "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: 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": ["cost_latency_tradeoffs", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aef09f514935d38e29124b17de287ed2e06cf95a"} +{"id": "train_44257", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3d74b66181885169efa652d72752f13a277f0c2"} +{"id": "train_44258", "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: 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": ["governance", "auditability", "reproducibility", "tooling"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afecf56309acd34a4bb91cad96abffbab8e31d42"} +{"id": "train_44259", "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: 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": ["auditability", "repo_scale_reasoning", "reproducibility", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c934ff367aa03ca7f31c520ccf0c1d9a7e3434b"} +{"id": "train_44260", "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: 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", "tests_are_truth", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c018a5561d0c945fb2a6276effa39d074b71650b"} +{"id": "train_44261", "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: 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", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c445379f6ef961a7b41fbb2685806a914d6f14a5"} +{"id": "train_44262", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d40763ffec00229e0977f661ccff8f9cbd0b35be"} +{"id": "train_44263", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bbab08a6958b7b73fe9bc1090f6610b9d6b3b41"} +{"id": "train_44264", "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: 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": ["cost_latency_tradeoffs", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fd25e0ac97842f1b52b47bcb68219ce3e6dc57e"} +{"id": "train_44265", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "823e0067ec3f303579eec8d68b1575461176a9df"} +{"id": "train_44266", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e8c0772a9436d01ce445490e1675a82d1d14dfa"} +{"id": "train_44267", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "179287bc14e9233c057ad5706f440ad22da7c744"} +{"id": "train_44268", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "security_gates", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c09a32538bd198624ae7166f1a682e4f7a261f97"} +{"id": "train_44269", "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: 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": ["reproducibility", "tooling", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "179c1ca6db73db61290d6cd414e943565b62dcf9"} +{"id": "train_44270", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e2341e2df02f7143282db0562a37bdd2f2f5bde"} +{"id": "train_44271", "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": ["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", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a40a49cc03bb4aeba7185f6c9d791035aef1b164"} +{"id": "train_44272", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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": ["evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e487bd7bac0b934bf5df5e7c6817d34e30e5cf4f"} +{"id": "train_44273", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fde192e4d99d36d24c049c63ef77bbd22466f3c6"} +{"id": "train_44274", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e0b0504eb96d010aa44c1d46ef290370e3b19cd"} +{"id": "train_44275", "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: 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", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cf8b3918b09a337d66d8d7475539cd5267c7002d"} +{"id": "train_44276", "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": ["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", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9f8dbfa52a94a333c8050a16c705c3fe6460ccd"} +{"id": "train_44277", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["cost_latency_tradeoffs", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b095922fbf5166761c30e1ff9c54ff72fbd33e71"} +{"id": "train_44278", "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: 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d8b4f463fe2d4a51d2ea0097b31d58ef37f004b"} +{"id": "train_44279", "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: 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", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5bb26c477fedd75e28506ef74d35b5e1e3f4926d"} +{"id": "train_44280", "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: 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": ["cost_latency_tradeoffs", "ci_integration", "documentation", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b742721083fad8e02cbbef4e5eb0b561afa09d15"} +{"id": "train_44281", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3beb7d557b342109b67cfaac5834c28e98a4c28"} +{"id": "train_44282", "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": ["documentation", "tests_are_truth", "auditability", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ddff4f14c989a00c25bac6aaf2243bfb6ef74bcc"} +{"id": "train_44283", "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": ["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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7aec61485d0f8d8ff638b5d56f88ae6666a75c02"} +{"id": "train_44284", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53fc02369e6f502e8c366a9fba86350e2e869f80"} +{"id": "train_44285", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "473525caf6d191e2167aa5e83046a1cc0f52c4ed"} +{"id": "train_44286", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb3e092e1c2560bf9dd9d0d5a7c9dd8b1fb3fd36"} +{"id": "train_44287", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f56382f9e29e34a14fb0b15e34afd6a29460b310"} +{"id": "train_44288", "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": ["tooling", "tests_are_truth", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed06e6468e432347eb4d979c2a735533fd01b2b8"} +{"id": "train_44289", "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: 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", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a38d3d78f873e656b6c5992b014b1f8394eb74db"} +{"id": "train_44290", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["documentation", "tests_are_truth", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "732da8f3942fdf8c89e00e44cb83344f487a9030"} +{"id": "train_44291", "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": ["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", "tests_are_truth"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b50fb4eadce1f8efbf0affbb83d0fc7e4a0a12b"} +{"id": "train_44292", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "31bc981788bf8e5abd6ca41834fe78aff76652dd"} +{"id": "train_44293", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6965227db9f47b5ae5a2ad843aa201195e75f8f"} +{"id": "train_44294", "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: 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": ["cost_latency_tradeoffs", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "870940ef96d3a3d726323dd757ee08d744a20acd"} +{"id": "train_44295", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "documentation", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ae1f700db0252932544f283350d3f29d837e3928"} +{"id": "train_44296", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87e2660383bad24f14c17f403321b3cce134d174"} +{"id": "train_44297", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tooling", "security_gates", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ef8c3e27501c58a1916e70ff498fdf9f54c0c912"} +{"id": "train_44298", "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: 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": ["reproducibility", "ci_integration", "governance", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "289f0d0f47ed34d0ac1d93c60ac8b040e2e260fb"} +{"id": "train_44299", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b84fcf366299bb231e4cbb57b791bb1ec4838fd4"} +{"id": "train_44300", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60d2b4438119fde49252621785258e271f72c236"} +{"id": "train_44301", "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: 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", "auditability", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1bdf9b082f2f8e89439c8fd3813b37f36c5ee2b"} +{"id": "train_44302", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9103a5bb5a288e7c557d94f62b920417fb37601"} +{"id": "train_44303", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "ci_integration", "governance"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3e4dfd4814d2fa16f9f6b81fc5bc4dd52023012f"} +{"id": "train_44304", "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: 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", "evaluation_metrics", "repo_scale_reasoning", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "083944f69aa0150b3051c1d8a3e5fd7466a17d0f"} +{"id": "train_44305", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "ci_integration", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a1db28b9be36c662ccd36133ecedba893f7d346"} +{"id": "train_44306", "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: 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": ["auditability", "repo_scale_reasoning", "reproducibility", "security_gates"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3cbd338784d048b8ddf23611cba2a2a3f842347"} +{"id": "train_44307", "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": ["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", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b76a148e23237b63d2e9d6088429c9a20b7068c8"} +{"id": "train_44308", "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: 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": ["evaluation_metrics", "security_gates", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f20435852eeec9baf1c61dc92e68c14a0c10e076"} +{"id": "train_44309", "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: 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": ["auditability", "tooling", "documentation", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8613941af692f1afd16d123425c6653ee1308b53"} +{"id": "train_44310", "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: 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", "tests_are_truth", "auditability", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85476f4c7612f2fd877fcaa4457c66306bce16c9"} +{"id": "train_44311", "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: 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": ["cost_latency_tradeoffs", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "125d8e50eb1516a7d33b6856260e60f1fefbad4e"} +{"id": "train_44312", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready 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", "tooling", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e129565ce285d33588c88d8f3342dca847173d4"} +{"id": "train_44313", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["governance", "auditability", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb3d9b3c6ccb2244447f5738a9f670464758e0a1"} +{"id": "train_44314", "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: 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": ["auditability", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "635d74d740cb990fb59074577953e28fa289b7c3"} +{"id": "train_44315", "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: 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", "governance", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7806c27867d3c7369fc12d2db69aa13f2f65d1bf"} +{"id": "train_44316", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "auditability", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e23989027bb01cdcb89620b13ca2ad61b850cd31"} +{"id": "train_44317", "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: 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": ["security_gates", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c8f5401e9d327aac096856adbe3ab1918995fa71"} +{"id": "train_44318", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "749ed6708588f0537cdb48d9878d9e6d8f837b22"} +{"id": "train_44319", "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: 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", "auditability", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f680c35b9744de4253b8e7bd81d483a1c60ef851"} +{"id": "train_44320", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "documentation", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "54baab375159da8b6ebb50f6f375218e705b52ea"} +{"id": "train_44321", "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": ["evaluation_metrics", "documentation", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4884693bf1b551f50b19b25ebd21d2c1e3efdce6"} +{"id": "train_44322", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "evaluation_metrics", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fa36afeb62e9deddd8cccf36ba0def7f317576c"} +{"id": "train_44323", "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: 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": ["reproducibility", "ci_integration", "documentation", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b14c3b0ae8a6d2ca07fc2055d67eefb63e936c37"} +{"id": "train_44324", "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: 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", "tooling", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ed4b91f70bf539922d25f3048b1fd2e5d493dfac"} +{"id": "train_44325", "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: 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": ["governance", "cost_latency_tradeoffs", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fde29419ed05ad893b624be1c9c01d4298fcaa69"} +{"id": "train_44326", "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": ["tests_are_truth", "reproducibility", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ead3831d6643d2c1fcaa1f84e57da28f7c7e50d8"} +{"id": "train_44327", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "881bb9b5a3d2894eeb7c7f145a3ed519a3d7bd60"} +{"id": "train_44328", "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": ["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", "governance", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6c6447f1b5d1d8e3d6934f8afc945a32793711c0"} +{"id": "train_44329", "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: 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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "security_gates", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ea6e205d8caa92a1badc685cb2c43c54c5d7356f"} +{"id": "train_44330", "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: 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", "ci_integration", "tooling", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ab7017642424f2eddf605348cf923fa4f1a10df"} +{"id": "train_44331", "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", "reproducibility", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff9abd1e2421ab37c7be2d5a8bec3eb8c82d941b"} +{"id": "train_44332", "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: 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", "ci_integration", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7411deb0a500b85515aae27ac5bb68b40fdeadcd"} +{"id": "train_44333", "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": ["repo_scale_reasoning", "ci_integration", "security_gates", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1c88172c0836998a3106a750fcf0c4737208fa8d"} +{"id": "train_44334", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab3de36494a4d82daf4d80ca33e4811e515eb43e"} +{"id": "train_44335", "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": ["governance", "ci_integration", "tooling", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7ad0e3a1b7152bfc9a6633ec4f52d4ca9d4a6e5d"} +{"id": "train_44336", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4c105bf55cd15406b1573fce73c0bbd2267184d2"} +{"id": "train_44337", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "017c57107f0db772db7fc927ea9aa64419e79d97"} +{"id": "train_44338", "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": ["tests_are_truth", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aeb7905e0ad4703dfb8894aa2d733ab71c562e46"} +{"id": "train_44339", "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: 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", "reproducibility", "tooling", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf0349e39214b50118aa581597e83e0a2f03de04"} +{"id": "train_44340", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a83aa3da5d8ff872c3c36495b6ba4a9d89424242"} +{"id": "train_44341", "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": ["cost_latency_tradeoffs", "ci_integration", "repo_scale_reasoning", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6776a724ec6a6758eb9d24ab1c003f511b8a2d66"} +{"id": "train_44342", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bda6141f23e1b8f90faab345e1a2375f7733417e"} +{"id": "train_44343", "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: 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": ["cost_latency_tradeoffs", "reproducibility", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e7a04f2ab28b84a7bf0f437dabe73e68bfb72d2"} +{"id": "train_44344", "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": ["tests_are_truth", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "528bb2d1a7ba2c25709825b19fd6d3a60ede2f2f"} +{"id": "train_44345", "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", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2d66175c49d98f2491e806b365fd77ff0662b24"} +{"id": "train_44346", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7c3b9a3bea7e6ea21b5187a7c4a0cece69dc108"} +{"id": "train_44347", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de41c6c0c669bcb4055200f085742a7c275a13e1"} +{"id": "train_44348", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "security_gates", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0717bbab8d779be708723acbad2e7ce1bdc87327"} +{"id": "train_44349", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e40e82949b17b905d58b01519bc31757b7c031c"} +{"id": "train_44350", "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: 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": ["evaluation_metrics", "auditability", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d560770526ebc82a24b835c828e09601537d0768"} +{"id": "train_44351", "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: 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": ["tests_are_truth", "tooling", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e690abdfb8e8e7506207eb3f6abdbc8018fb00d0"} +{"id": "train_44352", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "tests_are_truth", "documentation"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "07f821c8e4a7cbb74a79a34b02cc9ed67eacdbe7"} +{"id": "train_44353", "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: 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": ["governance", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0740d1fd0e9ad48a2206c229e86c0c48d4bbd4a9"} +{"id": "train_44354", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "documentation", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3a08f1dd50f3c377942a20314433fcfbabdb19f"} +{"id": "train_44355", "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: 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "26a78d74c64f4832a4f9b971afce4d2a9aff246f"} +{"id": "train_44356", "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": ["ci_integration", "documentation", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d420a7b05128d81da5dfa248ac6dcfae0bf7a308"} +{"id": "train_44357", "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: 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": ["security_gates", "auditability", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8fba8ec1374df2246a08ea11a3c11a81a5f0636b"} +{"id": "train_44358", "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: 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": ["tests_are_truth", "evaluation_metrics", "reproducibility", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e4e1b684927fff47867def3e7ee0d97fcd0b87b2"} +{"id": "train_44359", "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": ["security_gates", "documentation", "ci_integration", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "34542abf6919fcd14f45b88064120bd515b43750"} +{"id": "train_44360", "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": ["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", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58e36ef314bc7d3d57365693253ec3cb49349a1e"} +{"id": "train_44361", "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": ["tests_are_truth", "ci_integration", "reproducibility", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "964d7c9f360dde82d78ca2a2868ca3847cbd65cb"} +{"id": "train_44362", "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": ["governance", "tooling", "reproducibility", "auditability"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "988b6f058f6f5b142fc11750275c4b08e25ed6e7"} +{"id": "train_44363", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40073d5143979aa41b61c2e0c77403af17e243ea"} +{"id": "train_44364", "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: 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", "auditability", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1bd755b5869fb8e40277e854c2a810a283bde921"} +{"id": "train_44365", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60a76163b750e62569de6e40aa00de1dfb9d4042"} +{"id": "train_44366", "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": ["ci_integration", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c7f431d7ad246451e014e779fcaa963220d7c032"} +{"id": "train_44367", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "cost_latency_tradeoffs", "ci_integration", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "016a8df867549d7005d299c033ea5409b45c8de6"} +{"id": "train_44368", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["tests_are_truth", "auditability", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ca0920a8b1b2844c7da849d6a6c0cce544b077d"} +{"id": "train_44369", "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: 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", "security_gates", "ci_integration"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0488a109981a2be35ad5284c9580713a65563b3a"} +{"id": "train_44370", "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": ["auditability", "documentation", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfaf866577e7e73d0962ae0d0e03ff70bd5c9ac7"} +{"id": "train_44371", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "541a4826cc90079099a7006091c69cca757dc675"} +{"id": "train_44372", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["documentation", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91b22f68ea9214948e39b262402a5fd4cb89458b"} +{"id": "train_44373", "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": ["cost_latency_tradeoffs", "ci_integration", "evaluation_metrics", "auditability"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82a76af5afc4b041a30f6677a22a609b7622ac9a"} +{"id": "train_44374", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d17e818fdb0b14d991236076d1bdff3e7a531c9c"} +{"id": "train_44375", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f18c5554fc1b0dbd3a8cb6518c162def4da8b1b"} +{"id": "train_44376", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "evaluation_metrics", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d03ebc4f2f209516718874f0f02506d5cb2baaa"} +{"id": "train_44377", "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: 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", "auditability", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a8be63737121d56bff827c7e9ab85b142ab11d6"} +{"id": "train_44378", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "857a18985a94bb9ff8af8adc34a94465a62c649c"} +{"id": "train_44379", "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: 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", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3dbef86dce2d2ffc0fc0048e13248976828a07a"} +{"id": "train_44380", "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: 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", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cfdccd36f1ce483f05dd490788b0707a406e51a5"} +{"id": "train_44381", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "security_gates", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5e1781451418649c29a98a9ce14062098a24a460"} +{"id": "train_44382", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "auditability", "evaluation_metrics"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aa3e9e69d4a8ca5411c6ea2e0a3b110f5f4e1e33"} +{"id": "train_44383", "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": ["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", "documentation", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d939d9f766d1892a95d0b6f078168dbaec84c2b4"} +{"id": "train_44384", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "762bb4efe9105e3d8b751726b0ecad5a0bab352a"} +{"id": "train_44385", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d105a3973e1ff52fa275e7b86967d0fe35c180f"} +{"id": "train_44386", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "tests_are_truth", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b32a4379510ba1e8fcfe1e6aa50b19345efc1032"} +{"id": "train_44387", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["security_gates", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dda31e07805703a0d35df0a0fdc4d1b61449d6dd"} +{"id": "train_44388", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8da939793924b6d4292426c95d5f8b96502c669"} +{"id": "train_44389", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2902daa1b4dcc91436e83a8fbb3afb5e9de4866"} +{"id": "train_44390", "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: 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": ["security_gates", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e8338a247da74c90fb31648746a2f551bccbf25"} +{"id": "train_44391", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "15cf058ddce006d33438ad1c2ac9b179bbf4c73e"} +{"id": "train_44392", "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: 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", "repo_scale_reasoning", "documentation", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91e99d59d0096920942d42d7cac99e1a9aaab569"} +{"id": "train_44393", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tests_are_truth", "evaluation_metrics", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "80008ae3ee0e34f176c438e4f90f7ce6cbae003e"} +{"id": "train_44394", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "evaluation_metrics", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8f408a23d49edd20fe46dfd3db1294c7cb9ef2b5"} +{"id": "train_44395", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8842b35f24ef34e4728b46eff664c8fa4994e95"} +{"id": "train_44396", "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": ["cost_latency_tradeoffs", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0923080ed5073f3f5e2692a4c08cdda61585390b"} +{"id": "train_44397", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7858744b5023807e551a9dd4dd539f421fdf80a8"} +{"id": "train_44398", "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: 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": ["documentation", "reproducibility", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85ff536d3f5b2bbcfce3796b57596b5c50684016"} +{"id": "train_44399", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "ci_integration", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dfb1eb2294eab25a20ce8347058341fb3633f246"} +{"id": "train_44400", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "ci_integration", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e37d21244c01690b8c7382071718eccfa4ede756"} +{"id": "train_44401", "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: 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", "reproducibility", "tests_are_truth", "security_gates"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "85e141f509a30c9efd2db61d9732b7fe0d00fc9b"} +{"id": "train_44402", "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: 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": ["reproducibility", "governance", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8473986b891b667207d7a1df99dbe0daac5c5ec8"} +{"id": "train_44403", "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: 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", "auditability", "governance", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8e527d000a086397788c091682f1618527c8bfbc"} +{"id": "train_44404", "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: 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": ["ci_integration", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60ef3b22350031b5c844fb14e4a416ae50dd50ce"} +{"id": "train_44405", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab06bdb360d7a6a312ec3b8ba6dc9ddc5f9d00b6"} +{"id": "train_44406", "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": ["governance", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5b9b4e61e80f10413c88d26592b9fbfe50dc9092"} +{"id": "train_44407", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5c550ba6e26c1ae27c14fbb2d41c76d984851aff"} +{"id": "train_44408", "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": ["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", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39aa1bea7a45d8d150801e87b6b38426c4e2e0e5"} +{"id": "train_44409", "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": ["tests_are_truth", "repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "743e2bf74b7e14981e02382fb135c09c48db1a35"} +{"id": "train_44410", "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: 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", "evaluation_metrics", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44184bbd035c438c1991338d7a82719b05775e16"} +{"id": "train_44411", "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": ["evaluation_metrics", "security_gates", "ci_integration", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bd1fc3b2405e8204299b1a368af34332d39fa716"} +{"id": "train_44412", "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: 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": ["auditability", "documentation", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0630df47e7acbda6d0c3bc1ff76c14ec6ff69a11"} +{"id": "train_44413", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "security_gates", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d203be4b58d97d3b3b45f24b0598910e36d9b9e4"} +{"id": "train_44414", "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": ["tests_are_truth", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8e91cf8eb6a842034cf9055ea6576f490c8422c"} +{"id": "train_44415", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2a71e0774cd394a18455de56ea0e60ceb9f60145"} +{"id": "train_44416", "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": ["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", "tooling", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5d2af4ee04bc6b9b417c4c580b103643b40ea49e"} +{"id": "train_44417", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "015db741bcbfb6e4aab020966f5f5e585ca74774"} +{"id": "train_44418", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e6e8b9eca1604e770fa049d979d906d679a614e"} +{"id": "train_44419", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["reproducibility", "security_gates", "ci_integration", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6abb2a0a3be4e5eaa1a0bae1cdee4f91798dc6f4"} +{"id": "train_44420", "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": ["repo_scale_reasoning", "security_gates", "governance", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bcb258de2e6cd99b05540b06426cdc6758b37962"} +{"id": "train_44421", "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: 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95fdf1f3faa0f78ab7594b9ed19cb8f3fb9292c1"} +{"id": "train_44422", "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: 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", "ci_integration", "documentation", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c6c0b0fe8e0e5d78e2c133414c6a528a4a836e0"} +{"id": "train_44423", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7c34c275319ccc55a6b65dbb9edc7782258f1a8a"} +{"id": "train_44424", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2e8ba9c530c7c61832bfe184dd102d67307df96"} +{"id": "train_44425", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "evaluation_metrics", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "49a8b50428f0d6145c82e122e39d73d75adafd01"} +{"id": "train_44426", "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: 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": ["auditability", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a046b131ded2418077600d4495a08934be7e0bd9"} +{"id": "train_44427", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3359570406cadf2d49b252ae854e2ef24f6ffa2"} +{"id": "train_44428", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "governance", "evaluation_metrics"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "619f21c0be1270e959f25314587a27d455a9aa52"} +{"id": "train_44429", "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": ["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": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2edf0d259227290dac65b1c1145a62173bc36fea"} +{"id": "train_44430", "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": ["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", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "289b5880b6b22d660a170842a93560694eaa814d"} +{"id": "train_44431", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9b6a5f9c7d4c019a68688162777a2521177a970"} +{"id": "train_44432", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e68622844137a07f46e846e241b9f3f22c531c52"} +{"id": "train_44433", "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": ["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": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "384f650b4c0b7cf79bc8201078187fe5b3c061d1"} +{"id": "train_44434", "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: 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": ["reproducibility", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87e0363b144ae08a48476f3c549e0a84643d828d"} +{"id": "train_44435", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "evaluation_metrics", "reproducibility", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0d1500829c49eabf8236be6ec005cfadee979a3"} +{"id": "train_44436", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d6c85ff896b516b8dc9031626f9b8dddafa8a794"} +{"id": "train_44437", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "tests_are_truth", "auditability"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ef71baf0c9b2769f0de6c95fd1c0dacd3c551fb"} +{"id": "train_44438", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6ea43a8e69d07dd236bb7ee5332eac1ab8df799"} +{"id": "train_44439", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c14222e2d95ab5a53e0920ca5d8372680e4dfaf3"} +{"id": "train_44440", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8a6a9cc36ec6cf8ffc69156dc3ee2a7f6052154"} +{"id": "train_44441", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "92f1e200c1ba9e3c1523b19645e2229283ad691a"} +{"id": "train_44442", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "337f915f3daa47a25ec36620d50af97e5c19a4bb"} +{"id": "train_44443", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3ce037b1154fe5cfdc560a9fe94671ce9c007b56"} +{"id": "train_44444", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab56a03f4e2200672e8008d735a478e2c438d68a"} +{"id": "train_44445", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["reproducibility", "tests_are_truth", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d8cf61144cb04f34792e79c63110b7e6a773285d"} +{"id": "train_44446", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["tooling", "tests_are_truth", "auditability", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c98ac729f5403de13ee616a8a20bc06a4edb1657"} +{"id": "train_44447", "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": ["ci_integration", "documentation", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdc91c16369ad1af94b28adcd23e36dd3f625fde"} +{"id": "train_44448", "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: 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", "documentation", "auditability", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c1dfee390d5ca22564d31e5490bb6e1b625171e"} +{"id": "train_44449", "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": ["evaluation_metrics", "tooling", "ci_integration", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b259e04d0b1638648477cea1eff2abfc02cff300"} +{"id": "train_44450", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "auditability", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b6d9b11302c027bc7fe313dd7304d987c75fd33a"} +{"id": "train_44451", "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: 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", "tooling", "repo_scale_reasoning", "auditability"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8016a7a88a7d91cab273df39275fdba5a391934"} +{"id": "train_44452", "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": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "02f84503e746944627b680c53e59a15700f70018"} +{"id": "train_44453", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "repo_scale_reasoning", "auditability"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9e528dafeb58b56aae093551b7cf5e16046ad083"} +{"id": "train_44454", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65cbb13ca028df86c87b83b987dcf8f287122ff5"} +{"id": "train_44455", "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: 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": ["ci_integration", "governance", "tests_are_truth", "auditability"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f320814bf3154dfc60818b41000d088e80f7ce3"} +{"id": "train_44456", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8aee599c96601f7a812725fb84ab306c52163b70"} +{"id": "train_44457", "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: 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", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7bea60fd5d007e583b43400c4afbdb27f4ff8d86"} +{"id": "train_44458", "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": ["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", "tests_are_truth", "tooling"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68f630f8f6cb042281133e33f9a2774fcc9e5c12"} +{"id": "train_44459", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b745e48c80d930b5bca672fecc5c37369ca26de"} +{"id": "train_44460", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "af581c4fd37c5353943d43c3f25a2aac25b07b0e"} +{"id": "train_44461", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3778419dfb0b19aa322d35f823fd596b39ee36e"} +{"id": "train_44462", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "013cc6b95744bbad3a8a2717c5b49133b04c2b1b"} +{"id": "train_44463", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "documentation", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "82f1d237a9804af159e234a233a7246514278e21"} +{"id": "train_44464", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "375af0e27b9c2766945eedbc5b54fffe5ca0808a"} +{"id": "train_44465", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["ci_integration", "tooling", "reproducibility", "auditability"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5e39f659081a094fb9fd34c78bf4d9e4155cb6f"} +{"id": "train_44466", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "14086844191eebc6ca8bd33c4ddeb73556551a58"} +{"id": "train_44467", "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: 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", "repo_scale_reasoning", "ci_integration", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ad07bc2fc525048b29b6e420ffb79f2f72d706f"} +{"id": "train_44468", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "governance", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "962a4b42ec8be58f36ddd2315e41439738438673"} +{"id": "train_44469", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00d01ba7d54f406f53131741d61bded26c6f4212"} +{"id": "train_44470", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "governance", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a52d4defef28a8df8c1290ccfb4753290bed862d"} +{"id": "train_44471", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "358482e25f7ecebc223a40e1f803f09bf9972f58"} +{"id": "train_44472", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56b404699850c3af148c0cea2c90818d9f404220"} +{"id": "train_44473", "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: 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": ["repo_scale_reasoning", "cost_latency_tradeoffs", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d76e3f1ac7130f736644d3167e4480e464ae39f7"} +{"id": "train_44474", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "reproducibility", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1d034e5584b8a75000abf3cc00b0a1257ae5d612"} +{"id": "train_44475", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "cost_latency_tradeoffs", "governance", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c741d1cf256f209fa9d720464291acbc4ff88ce3"} +{"id": "train_44476", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["reproducibility", "auditability", "governance", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecab2ed9410362ac4ee10f3a5f81a180d95b318d"} +{"id": "train_44477", "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: 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": ["cost_latency_tradeoffs", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2226a5f62aeae03f32d569619a539033f30716c"} +{"id": "train_44478", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8840af94a7da1b7622c582da7473bb566ec1cb1e"} +{"id": "train_44479", "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: 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", "tests_are_truth", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5ee93d9e1b35f5571293a1f8333dabd2413aa870"} +{"id": "train_44480", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "governance", "reproducibility"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cbb8566331002f8512de81b921d0b9fce8abdd0"} +{"id": "train_44481", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "tooling", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "afb05e99f5e9b7badb7a14993db076eb8bd1d622"} +{"id": "train_44482", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "documentation", "governance"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94c23b67986b37d56e0537ec3a5f4a519f808fa0"} +{"id": "train_44483", "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: 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", "security_gates", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c4656c710ceac33b9c5e0f2d3404c05e787cc3c"} +{"id": "train_44484", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29969df0950ea99dc3f31856ea759f569cba8894"} +{"id": "train_44485", "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", "tooling", "governance", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "590109a03440fe0858dc4580b3d874739b9461d2"} +{"id": "train_44486", "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": ["tooling", "reproducibility", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cefd391273ffa4df92931fdaefef3a69b0733f8"} +{"id": "train_44487", "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: 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": ["reproducibility", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d25cf2e974e1c9d92caadc961f66ea6555758f6"} +{"id": "train_44488", "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: 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": ["security_gates", "ci_integration", "governance", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7411deb0a500b85515aae27ac5bb68b40fdeadcd"} +{"id": "train_44489", "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: 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", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de5b11ad64834eda1cffed307b1e0bcde60acb86"} +{"id": "train_44490", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "ci_integration", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "982b08c182f3764beb5179a6841991851096b61b"} +{"id": "train_44491", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55c7ef9fe42605c074dafcde9753e665964ce301"} +{"id": "train_44492", "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": ["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", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "04cffffa32f25b726a1cd0d80785899dbb9a87af"} +{"id": "train_44493", "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: 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": ["reproducibility", "tests_are_truth", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "677a43f6b7842941d715f4a2ff995ca7e04c557e"} +{"id": "train_44494", "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": ["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", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d57c8c65901ec53d8e0ed53ffa764ebd32cc8513"} +{"id": "train_44495", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aaeff21e7f34ac625c7b036eb2107d6b1399c69b"} +{"id": "train_44496", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c602db113e5ff6725ad62e3279150edc6615f99"} +{"id": "train_44497", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40d257f899a5371f0c93114dcee5d1d8211fa108"} +{"id": "train_44498", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "security_gates", "tooling", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fd455646e2281eaebd2be25f3e7875e1546d51a0"} +{"id": "train_44499", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70b11b84be0213249f39ff2d20b2e07372c4a91f"} +{"id": "train_44500", "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", "tests_are_truth", "evaluation_metrics", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fa8b48f5408f68e851662a7e46b0c48a66b46157"} +{"id": "train_44501", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "governance", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44edcca5bdfe6dc0e2f4f9692f89ca58e11764f9"} +{"id": "train_44502", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["tests_are_truth", "governance", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5acd8ccc6190d420f139704c62226622694bba0"} +{"id": "train_44503", "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": ["governance", "repo_scale_reasoning", "auditability", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "289b5880b6b22d660a170842a93560694eaa814d"} +{"id": "train_44504", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "864310f43972cc6ffaa280f132e2844b9c045495"} +{"id": "train_44505", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "security_gates", "tooling"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "11b4ef8944e905960f6a7d1c99b0ea9c110f6cfa"} +{"id": "train_44506", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5adc8701953ceb7831b7334e9dcc0dc1a69a94da"} +{"id": "train_44507", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e333521ecf5b334cf4e9a133d0f020b0ae123052"} +{"id": "train_44508", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "auditability", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0448560e6c9e469c565240354633d6e8a6614cb0"} +{"id": "train_44509", "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: 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": ["reproducibility", "ci_integration", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b8c535d2afecf6071dd5e8ee9aa6ec63deeb29f"} +{"id": "train_44510", "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": ["tooling", "governance", "documentation", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "97f979f9517da9b062e1eacf2fb865701573ca7f"} +{"id": "train_44511", "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: 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", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "46dc563a87ae6761b986a82142c304fca2157be2"} +{"id": "train_44512", "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: 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", "repo_scale_reasoning", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bebd13c45819a05f2be28d6ac219a8711e100bfd"} +{"id": "train_44513", "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: 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", "tests_are_truth", "auditability", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fac73e2a18e3b2f05daa589891ceee40be0db753"} +{"id": "train_44514", "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: 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", "tests_are_truth", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e079ae03aa58ce292d0a98f25f4a360101f62589"} +{"id": "train_44515", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "ci_integration", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6c145cef6418ee8f71c83374951789b20d8f380"} +{"id": "train_44516", "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: 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": ["evaluation_metrics", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bce3b7a84c111edcf2f032ca060f614855b5d4e0"} +{"id": "train_44517", "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": ["documentation", "ci_integration", "tooling", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5cba41f1ae4750d8768cd0619dd924dc13f10ef"} +{"id": "train_44518", "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: 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", "auditability", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b74a44bbbe0fe1c29a018d2eb343b04cecc7e94f"} +{"id": "train_44519", "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: 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", "governance", "reproducibility", "tooling"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "963cd2f3bfdac6ea45c9ada9aa1ce1dbd33d2ed5"} +{"id": "train_44520", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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": ["security_gates", "tests_are_truth", "tooling", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b26d08b331872df5b714737c4e73dea3c50cf5a3"} +{"id": "train_44521", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "649e198574286926963e28a031160f7a29452429"} +{"id": "train_44522", "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: 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": ["ci_integration", "tooling", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f4c14cc0fc365959d7a90afa572002c492af50eb"} +{"id": "train_44523", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "security_gates", "governance"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fbc4765fa3c2153d74d53cdd5f4565410809356"} +{"id": "train_44524", "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": ["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", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "788c1d3d763e3201bfdaf191d6fef9364e9bbace"} +{"id": "train_44525", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["security_gates", "ci_integration", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a944281425809492f1cf167373156382b931abe6"} +{"id": "train_44526", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b67733acff748ed982f7122f2c9cdf4ac08fe400"} +{"id": "train_44527", "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": ["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", "reproducibility"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "58db8de21105f42b6f7a69ba7e2253d275683f38"} +{"id": "train_44528", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2d8482ff0460f081e5222389478714e656e7fbb7"} +{"id": "train_44529", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17ae965549853393c49fa8beffa921df67f5895b"} +{"id": "train_44530", "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: 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", "tests_are_truth", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7955a7ecd5e6698cee9512dea0e1fd9f158883e1"} +{"id": "train_44531", "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: 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", "reproducibility", "ci_integration", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1eba24d9a51752d5d85c438bcf65b4cc4881402"} +{"id": "train_44532", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b32617d58afe2471ba508c2c4c42cbc7ae84521d"} +{"id": "train_44533", "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", "documentation", "tests_are_truth", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfc6f4aa4ba246a63b5206e06a1465f51744a6af"} +{"id": "train_44534", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4fd674fae3e0066e7a2dbe25510ea1a8b679154"} +{"id": "train_44535", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9a51d64c65c1505b8787e65ec33e4a513f8cb51b"} +{"id": "train_44536", "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: 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", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4293ecfccd76480f0a0d53170bd73da9a603b31"} +{"id": "train_44537", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "security_gates", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e3147673bc89f032149880601e00246c691a4847"} +{"id": "train_44538", "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: 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", "auditability", "security_gates", "ci_integration"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5aad670956d200b3f3588368fc6096ab3d6ba18c"} +{"id": "train_44539", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d30872d4c028bd77f2debf862c0d14d4d95ed7e"} +{"id": "train_44540", "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": ["cost_latency_tradeoffs", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "12d10c0ca91dd1e073e18d87446eaf553c9b1387"} +{"id": "train_44541", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["tooling", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27d328cdfba109b1eeb8efbf4dc565b177bd0fc6"} +{"id": "train_44542", "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": ["tests_are_truth", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "38ea56ab28bf11fe897511252f8bd7dd423edc17"} +{"id": "train_44543", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "797dde6a9bbb0379abbea794797044c2bbd0b3e8"} +{"id": "train_44544", "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: 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", "security_gates", "reproducibility", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53201ddb02d1bcadd1c522035cb9f92066226552"} +{"id": "train_44545", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f962232e2a76e250c576a1fdd9c72bd72337550"} +{"id": "train_44546", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2f00e56169d6a9c40c564f6f4672c831ae62aee"} +{"id": "train_44547", "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": ["governance", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cdc173b23a1b83f706773cfe3a9dfe82d9684ec5"} +{"id": "train_44548", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19009fdb6f1b59f34c5eda5d4931a02e54b6499e"} +{"id": "train_44549", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "documentation", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a0c3df38e8d3ffddfb9c4d5a0cadfb9ad5530a03"} +{"id": "train_44550", "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", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "425a88cb10f2d4818159eb7349d208dc3d9340bb"} +{"id": "train_44551", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "security_gates", "tests_are_truth", "reproducibility"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f496a7a88e6055717f6aed9b2cbf4b864a3d9e6"} +{"id": "train_44552", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dea343aabcdd00d48be67c2e2f8969648babfe43"} +{"id": "train_44553", "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: 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": ["governance", "security_gates", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8750b697a845315552380f221b7ae1cb0b6bf63"} +{"id": "train_44554", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "governance", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "74950df59114b75a8273ddf9b1686cf9e1316486"} +{"id": "train_44555", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "governance", "documentation"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36baa7a3a058292d82a1c2175384b4af53b9b6a0"} +{"id": "train_44556", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cc86265bcfb981a2f7f74f6a3862abf4bd06eca5"} +{"id": "train_44557", "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: 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", "security_gates", "governance", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28b8c7a8d41505f8d36c04323ecdda1b8d6d7c57"} +{"id": "train_44558", "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": ["tests_are_truth", "governance", "security_gates", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "795698d3569748dea7a71f536cfd28f73d5639e9"} +{"id": "train_44559", "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": ["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", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "072ceccf970e7af468a6fbafefa9a50665eceaf3"} +{"id": "train_44560", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "governance"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9f61fe498703f4c7ddc2d9550a09beef690188c6"} +{"id": "train_44561", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccdc608073d1fe9242c7694f0ad47c85d9143845"} +{"id": "train_44562", "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: 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": ["cost_latency_tradeoffs", "evaluation_metrics", "security_gates", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "924c0507fefd78e67bbeb637fe30641f2901735e"} +{"id": "train_44563", "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: 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", "security_gates", "repo_scale_reasoning", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83e74998f54d30073e2cd7fd90ddd17e93eb396c"} +{"id": "train_44564", "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: 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", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f8967ac7f1af168c25049e9f337942e46fa0c3c"} +{"id": "train_44565", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "76195036db6e14ea133f89df86075d1a3d9d68d5"} +{"id": "train_44566", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f56382f9e29e34a14fb0b15e34afd6a29460b310"} +{"id": "train_44567", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "auditability", "governance"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3fd211de501d936b3a431f682bb313041c0374c3"} +{"id": "train_44568", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94fdebf6e39d89ade1d8ebc297f6d3fb48dd3d47"} +{"id": "train_44569", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "457566581d8e43c5f9459acaff7dcb72eabb2347"} +{"id": "train_44570", "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": ["auditability", "cost_latency_tradeoffs", "tests_are_truth", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2abb516b785d0a08d483043a9606fe7983362a4"} +{"id": "train_44571", "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": ["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", "ci_integration", "auditability"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0eed7df232b1e461316c94530635a0ed6986dde5"} +{"id": "train_44572", "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: 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", "tooling", "auditability", "evaluation_metrics"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de422b335530323856876a93db3f5554bc54a160"} +{"id": "train_44573", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "reproducibility", "governance"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7b3b8a90e5e137f9a82e04eb1401842784b09cb"} +{"id": "train_44574", "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": ["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", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2c0cb02d2979ef213dd4ad360239e4eb79801b74"} +{"id": "train_44575", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3f14f5b4459ce7f57d43cd3b078d227e12d6929d"} +{"id": "train_44576", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0f304ce75a9327d3184c6d5ae3e1c97f8cca1e1d"} +{"id": "train_44577", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdba311d7343e5d7701e9de961a3297908ea119d"} +{"id": "train_44578", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "45816a2c8d0fbfa070b537d2b5030fd1be773570"} +{"id": "train_44579", "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": ["ci_integration", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5891b30153211125ea811bb0460c83c6b62935ee"} +{"id": "train_44580", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db71a26db7f47ffd4e74a1893859906cf1d6766e"} +{"id": "train_44581", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "documentation", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1012d5660e6a9c325d6e7d1ce145be9a4d7eead"} +{"id": "train_44582", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "009b0e745047fbde89e5c0c5eea892babf0b4e93"} +{"id": "train_44583", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "cost_latency_tradeoffs", "evaluation_metrics", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a483e6afdecdb4de5934f43ceb68749eeec3350"} +{"id": "train_44584", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "33e14e93eac888c009df11487ffeb51bd98ca281"} +{"id": "train_44585", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "81667b547f60eaf54634a4671061f651e407be5d"} +{"id": "train_44586", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "db01d5574ceaef5cfe227783d21f14eb16a5d056"} +{"id": "train_44587", "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: 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", "ci_integration", "reproducibility", "tests_are_truth"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8f1586a9c0a7e3db810bcdde3dc74a4f45220b6"} +{"id": "train_44588", "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": ["reproducibility", "security_gates", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c1c2040af1f2e8b03d313ec64c4be68d60bc2056"} +{"id": "train_44589", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "security_gates", "auditability"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dec9f1fa33d2ef11d0c26a173923c16a63bf3fd"} +{"id": "train_44590", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7853b77aa863d76644b4cc36a8eef5565e0eacbc"} +{"id": "train_44591", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "security_gates", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3c93e3aaddd738ed458ea4677b908b7cc4ed00b5"} +{"id": "train_44592", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "documentation", "tooling"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3229d8917e4047c74d14afdbfabcd8549afa0786"} +{"id": "train_44593", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready 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", "governance", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "540eba8f742735870cb183c23e34215535507d1a"} +{"id": "train_44594", "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: 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", "repo_scale_reasoning", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a3ee21fc98941bcf7f2d286b070d73da8c58d8ef"} +{"id": "train_44595", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a17157f3983a2760ebf7b6c1efcfb62ae4d3185"} +{"id": "train_44596", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f5bda5832f8a7529af123f2d95c16e326198571e"} +{"id": "train_44597", "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": ["security_gates", "documentation", "reproducibility", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ac1f9abdf028f35042561f4c553d8c657d480bd3"} +{"id": "train_44598", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "repo_scale_reasoning", "governance", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f2594c391bab5c272c8ecf6ca51033b514d398c5"} +{"id": "train_44599", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["governance", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c514161a197ce05a31853cc883704cfd7d394509"} +{"id": "train_44600", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "tooling", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daa6fc42a7b7303c918f616bbd3ade7841b99b63"} +{"id": "train_44601", "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": ["auditability", "tooling", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "be392eba4a20b071198ead54fb3d0bfa6c90df05"} +{"id": "train_44602", "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": ["auditability", "tooling", "documentation", "tests_are_truth"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53ccc119b3cde2e71cc00feaec48cb91d8d870c2"} +{"id": "train_44603", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "documentation"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e1071dbdd2e025b55f58640e36a009ea463ea64"} +{"id": "train_44604", "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": ["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", "documentation", "security_gates"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5f334bd1924227896206c15d0abe6ee240ed0804"} +{"id": "train_44605", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "aeac424e5ac2082821f1b88cce26585903dcf9bb"} +{"id": "train_44606", "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: 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": ["cost_latency_tradeoffs", "documentation", "security_gates", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4123cfa5c26525e847bddbaa866e0083a615af1b"} +{"id": "train_44607", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4d70b509df7db22d2a7f32eddf1fcfd94a9ed9d4"} +{"id": "train_44608", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "documentation", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75e85904c93e472afa3be89b01bb715886bb5903"} +{"id": "train_44609", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36a66aaacc10516a493f5b2dd9026a52905268c6"} +{"id": "train_44610", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "auditability", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18e40a686198aa715b6c83bcf859b4648dc6731f"} +{"id": "train_44611", "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": ["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", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d5840a595c29379955af6d79fcd45941d7d67fb3"} +{"id": "train_44612", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "659eeb3d2873689b337b8818034b882a4d0e9bf0"} +{"id": "train_44613", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20b84484fb4f8de45443f72004017c6bda7e1c26"} +{"id": "train_44614", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "60b443e731e3f36d914689678b237be77328d9da"} +{"id": "train_44615", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65766eb1c0b48ab8417012c279f50a71e948b57c"} +{"id": "train_44616", "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: 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", "documentation", "governance", "auditability"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bdcd57c81c1ce51e544611bf5dff2395f9dce990"} +{"id": "train_44617", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "caf0321135f59059795dc86ce3f4f9be323fcc6b"} +{"id": "train_44618", "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: 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": ["documentation", "governance", "reproducibility", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e9c34559914d40e232b857acee4d0c1f1465c7fb"} +{"id": "train_44619", "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": ["tests_are_truth", "cost_latency_tradeoffs", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "de2aca3aba0cf05f05db71313928d34fd5143ccb"} +{"id": "train_44620", "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: 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", "tooling", "reproducibility", "governance"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "36da77c431cb0e78c9d7957f2aca21eb9a65ac1a"} +{"id": "train_44621", "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: 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": ["ci_integration", "security_gates", "governance", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "448ae4ac8fd93f403e29d603c0ac27ad22d7bfe2"} +{"id": "train_44622", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "repo_scale_reasoning", "auditability", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb769ad27bbf7b27b918f24072815ae69924ca63"} +{"id": "train_44623", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tests_are_truth", "auditability"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "00d01ba7d54f406f53131741d61bded26c6f4212"} +{"id": "train_44624", "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: 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c19e5e0399cd3b9decac3bef673a6ce3d6f001af"} +{"id": "train_44625", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "evaluation_metrics", "reproducibility", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44da5a462ef4f281f76a3ac7dbef055edac39561"} +{"id": "train_44626", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "repo_scale_reasoning", "security_gates"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a2d44881953707a8c83565cb826b2b7b8659eb6b"} +{"id": "train_44627", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "security_gates", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47b6aa581dea2f542ab0d7b4a74508c8910a9c5e"} +{"id": "train_44628", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d3b6c1d9a468552f70beae90976c4a4b688472a"} +{"id": "train_44629", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "151a46eae36768b5b1a81c7d7309f253e7678c38"} +{"id": "train_44630", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f3858004cd128df0fcc8046aa5c4ee54f19004fb"} +{"id": "train_44631", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["evaluation_metrics", "cost_latency_tradeoffs", "tooling", "reproducibility"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "788a83f13dd03a183ffcfde8d8859833306a97f1"} +{"id": "train_44632", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "tests_are_truth", "documentation"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71a50d1b52c993475fb517dcb7cfe1e4f91ae876"} +{"id": "train_44633", "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: 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": ["documentation", "tooling", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52c8079e530a1246eb4e1405804625110ff991d9"} +{"id": "train_44634", "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: 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", "security_gates", "documentation"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "815ed099955f8a084b1228b54a0676a37bf6f7b9"} +{"id": "train_44635", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "88cd211d7cefa9322cf3727027eb1a13d3f8ef13"} +{"id": "train_44636", "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": ["reproducibility", "tests_are_truth", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b259e04d0b1638648477cea1eff2abfc02cff300"} +{"id": "train_44637", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0d6f2610b38757b59ad3d1a4ae9af1b735c36668"} +{"id": "train_44638", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "089c152d02c946e67c49f17a19c4f502f5d0523d"} +{"id": "train_44639", "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": ["tests_are_truth", "repo_scale_reasoning", "governance", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "780456c6eb0b18f50df8e5e0aaaecc2601550dd4"} +{"id": "train_44640", "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: 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", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3becc07cf86fb5146564aa265b042ccee16978e4"} +{"id": "train_44641", "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: 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": ["governance", "tooling", "tests_are_truth", "documentation"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1563357e66c2480dae2522844ccaed364f69c1d0"} +{"id": "train_44642", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "ci_integration", "reproducibility"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "734f9c378a0857223050843810823e170ce9d27d"} +{"id": "train_44643", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["ci_integration", "governance", "security_gates", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9106ab2e5f217f3446f99845a4bf34ebaa15d0f2"} +{"id": "train_44644", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f5136c7650abd507dad177eaa7be60d656c25a2"} +{"id": "train_44645", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "security_gates", "repo_scale_reasoning", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ac755972c953c4fac27e9727fbcb039a7eab3a3"} +{"id": "train_44646", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "auditability", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52e80b695a5bfb3d88ea89e59d5b5e0294444671"} +{"id": "train_44647", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc18116889b69656cd9abfe4f2acca8f6d454601"} +{"id": "train_44648", "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": ["security_gates", "tests_are_truth", "ci_integration", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "59dcca3e66246ac8ac42c5b92b4aec5632ff9f2a"} +{"id": "train_44649", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "tests_are_truth", "reproducibility", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39a138030ee4ac0d3464c15849128c20f830eade"} +{"id": "train_44650", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "41a04286dc87d29842a6ee17cde4d3829b98af5b"} +{"id": "train_44651", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4dfacf8c30d7336da7f3a01593745f5dbad5ab15"} +{"id": "train_44652", "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: 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": ["reproducibility", "tests_are_truth", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f680c35b9744de4253b8e7bd81d483a1c60ef851"} +{"id": "train_44653", "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: 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": ["governance", "auditability", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c5a7085bbd8e0063738302792f2b0d6b83e9a50c"} +{"id": "train_44654", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da672554539ba693659911120c9f2accd6abb711"} +{"id": "train_44655", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "28d08a3295bdebe1cb2c29f9db62470426cf62ce"} +{"id": "train_44656", "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: 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": ["evaluation_metrics", "tooling", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "22a2eb5bda6b24871ce23b2f25d0f4d585eaef00"} +{"id": "train_44657", "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: 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", "tests_are_truth", "repo_scale_reasoning", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "56b85187047e61756a1e97302b4a02d83b5d8513"} +{"id": "train_44658", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8ee3e7e5c150559df1ddf58e232fd17ae9688964"} +{"id": "train_44659", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4bba865387cf8093ee43a916f365844eaec7bef"} +{"id": "train_44660", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["reproducibility", "evaluation_metrics", "documentation", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "62448bb0a73dda026db2da8d1bc83835bf847385"} +{"id": "train_44661", "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": ["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", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c4f229db37e1afb65304e191e73068599ba3dacd"} +{"id": "train_44662", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "ci_integration", "governance", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86f8b1fb7c3428738ae7d5b11bb46490a34bf0ef"} +{"id": "train_44663", "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": ["reproducibility", "governance", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ca50b7352d18fc5cb6eb519a3b0601249894b5cb"} +{"id": "train_44664", "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: 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", "ci_integration", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "035ccdc979993c57fb21dd267d7171469f53b0ad"} +{"id": "train_44665", "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", "governance", "documentation"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "91722486214f4e38efe08c988f1757bae46711dc"} +{"id": "train_44666", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "security_gates", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fc4d6408c86b4598704e175595663bb423020a54"} +{"id": "train_44667", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "governance", "auditability"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "293fd1f09857e73612b114fc7f76789899057c11"} +{"id": "train_44668", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d73c41fa9196d589999cef076b286c89f61c3366"} +{"id": "train_44669", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "repo_scale_reasoning", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9c9505a3facd3120bf4890c017fb6c9de2a23c4a"} +{"id": "train_44670", "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": ["tests_are_truth", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ccbe01e1ac8bf984f15b7c05348e19a73c4ed51c"} +{"id": "train_44671", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bc5b30ac9e03bb5953c096b2df0774e83794d038"} +{"id": "train_44672", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "security_gates", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "17b759c8f43062cb20b086efee53f63da9867f83"} +{"id": "train_44673", "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: 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": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "37bff38f43bfe58403adce4f8d2a6cda06b7c1e2"} +{"id": "train_44674", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["ci_integration", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e0c47d45d15ec40775403edf6e899b088aab687e"} +{"id": "train_44675", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "auditability", "documentation"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a39e5200be626df62717a7dff51c375987b1c6bd"} +{"id": "train_44676", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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": ["documentation", "ci_integration", "tooling", "reproducibility"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "255b185fc9f3720c7e950f719c173740a940ce70"} +{"id": "train_44677", "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: 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": ["tests_are_truth", "security_gates", "auditability", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a5322f219a06777b4e2bb1fc0bc8a3fda63a811"} +{"id": "train_44678", "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: 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": ["auditability", "reproducibility", "documentation", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3b64ad5b01977880f499992034e01bc7d87ed9d9"} +{"id": "train_44679", "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": ["security_gates", "tests_are_truth", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0690c7c6aba656b7c6316df3ac973a5ad3fb1ec3"} +{"id": "train_44680", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a39194e707a86a1e0ef46249771dcae44bbf3cc2"} +{"id": "train_44681", "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": ["tooling", "security_gates", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cab07b1db1382b0a1b75c5a9517db92e3b5fc59"} +{"id": "train_44682", "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: 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": ["ci_integration", "cost_latency_tradeoffs", "governance", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8438c4e0f99838b9b8e4a1e1c9740b255dc30d65"} +{"id": "train_44683", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "governance", "documentation"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebf5f532383cf882e5dbd7100153384dfd7e02a0"} +{"id": "train_44684", "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: 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", "documentation", "security_gates", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44e1f2bba462341195e2bb3ae673667d64abee6d"} +{"id": "train_44685", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8094b5902d5bda6aad4058234551aa50065b55b"} +{"id": "train_44686", "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: 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": ["tests_are_truth", "security_gates", "reproducibility", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84f6fccd5481ac972203c8c20a0156e9442f3b19"} +{"id": "train_44687", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "reproducibility", "security_gates", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bda778d7be2f5aa8095163bb5cca06f2a83b60e5"} +{"id": "train_44688", "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: 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": ["cost_latency_tradeoffs", "tests_are_truth", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "809f8ee0bb01e33b99a236fd82ead16e237df054"} +{"id": "train_44689", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "cost_latency_tradeoffs", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c15c1216978dfea0fc55f49a0aa7aa20b26c1943"} +{"id": "train_44690", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a4c8b6f840dc6f2c1c7ee92df1db742c2919bc69"} +{"id": "train_44691", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "tooling", "tests_are_truth", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7759306d913f6b6779080ce60664ac132d359941"} +{"id": "train_44692", "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": ["reproducibility", "security_gates", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2f85a69bee67c048405e37c6475f1fd07da85fc"} +{"id": "train_44693", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["documentation", "reproducibility", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "efac26b08c24271d59a0e35e87df8f8effb5b5db"} +{"id": "train_44694", "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": ["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", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3a6b83fac7e9ad52a4d97a6f53f082ec7301c9c8"} +{"id": "train_44695", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2f00e56169d6a9c40c564f6f4672c831ae62aee"} +{"id": "train_44696", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "evaluation_metrics", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "156e098190817b90aa0bd1a1e7c0ce38dc690c3a"} +{"id": "train_44697", "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": ["cost_latency_tradeoffs", "ci_integration", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b639c5452a90e7b11e47b50989f186eb7db8f68a"} +{"id": "train_44698", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tooling", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4e93d78c5e3278433b518ba7691b4fe9acff065"} +{"id": "train_44699", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "ci_integration", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b028537734256931af141a2a540918edea3021f9"} +{"id": "train_44700", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "auditability", "tests_are_truth"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7fdf673b40039f38051b78437c41507475d17854"} +{"id": "train_44701", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e927a8de744c50812e1f29d203cd8605d9e702c5"} +{"id": "train_44702", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "tests_are_truth", "evaluation_metrics", "governance"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "473403759eb38b3118d0d8e5dff30bcb7c3ced3d"} +{"id": "train_44703", "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: 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", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86cad0b0cfb8f71735785306ceaf48394e3e2ed5"} +{"id": "train_44704", "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: 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": ["cost_latency_tradeoffs", "documentation", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "775c528dcec5838ce47db7a33ba8c70ec5efacf0"} +{"id": "train_44705", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["governance", "security_gates", "tooling", "tests_are_truth"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2e4c4953e4c2ac7282b284ce434beb85cf61c3b"} +{"id": "train_44706", "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": ["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", "repo_scale_reasoning", "documentation"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71aa36d01fd476d061e65a302346673a07f3495a"} +{"id": "train_44707", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "tooling", "security_gates"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e292a0e87216cdc608e6b0e088e3411b7b2745d"} +{"id": "train_44708", "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: 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": ["evaluation_metrics", "security_gates", "governance", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05dc80ff22ae0d3c2c4db4657f7d1cbb4156e13a"} +{"id": "train_44709", "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": ["governance", "auditability", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1158260df36c7c2462e7bf2f508cbcedc1f1f80"} +{"id": "train_44710", "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": ["governance", "security_gates", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "038c75543f531aef0538b4af74ff4e4a1d94af0f"} +{"id": "train_44711", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7afdf89492526efbbcbefe940760b5d9b54c36e8"} +{"id": "train_44712", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0cc4a7f1604b0b5329fd07f1bb3b240759353ccd"} +{"id": "train_44713", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "tooling", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e8ab5904e8779fabaec7d5fc71d54381d58d737d"} +{"id": "train_44714", "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: 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", "tests_are_truth", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ff3312b0b5d3c24eabcf76fccf09242f1912ae9"} +{"id": "train_44715", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "19a4a738e7ea58e92a65719a9e7312fdd92968cb"} +{"id": "train_44716", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "security_gates", "evaluation_metrics", "auditability"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c01c64c7d004d93c47b9ec5bdfb7052eb07184e2"} +{"id": "train_44717", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "evaluation_metrics", "cost_latency_tradeoffs", "ci_integration"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f98444a4f4086336fa30496bbc4714c539455fff"} +{"id": "train_44718", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7d6fd9697f8646ca2fd3561e4ea73397af1e8cce"} +{"id": "train_44719", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "cost_latency_tradeoffs", "reproducibility"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c98ac729f5403de13ee616a8a20bc06a4edb1657"} +{"id": "train_44720", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "reproducibility", "security_gates", "governance"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "47924e2c503da0b3e683903f94d08ba80ffe97f3"} +{"id": "train_44721", "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: 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", "cost_latency_tradeoffs", "auditability", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f6e4dd51d4268f5c6c49e5a6669439c10574999"} +{"id": "train_44722", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "73101921169c5ce7b7673407e4e62e000d5db79c"} +{"id": "train_44723", "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: 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", "repo_scale_reasoning", "evaluation_metrics", "tooling"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1a1944ce4e90b6102cb5298900ecd670022c9a34"} +{"id": "train_44724", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["tests_are_truth", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "831307fe7338affff352a194811d29ac9183a032"} +{"id": "train_44725", "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": [], "metadata": {"target_language": "SQL", "developer_needs": ["governance", "security_gates", "ci_integration", "reproducibility"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a03a09131207a259825c6b5571702a9d8881ba78"} +{"id": "train_44726", "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": ["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", "reproducibility", "security_gates"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b59792a66f35d80d6f7e07bd539b5ae7f1ee9d67"} +{"id": "train_44727", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a570677acad9ea19e6275d9ae81a30818a7ba091"} +{"id": "train_44728", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "security_gates", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57ff5dcf05c45aa47373bf7ba1711247b0e07ab6"} +{"id": "train_44729", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2bba3fc4ae467e6b048b1f7d7ca6a1f4380c7255"} +{"id": "train_44730", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "42018f75da4e982c9080df3f5dc8378c85df1eae"} +{"id": "train_44731", "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": ["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", "documentation", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "48a93f43173578ff3d388e2cf962d6f95ce738eb"} +{"id": "train_44732", "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: 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": ["repo_scale_reasoning", "evaluation_metrics", "auditability", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "102c9f6e44d378faf9e5e9f515813f73defefd45"} +{"id": "train_44733", "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: 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", "tooling", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "823173ab02ec584be033472a6449e1fb8c01ebbf"} +{"id": "train_44734", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6719b931bc24cbdbb184c2d2209c4af014f979a"} +{"id": "train_44735", "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": ["evaluation_metrics", "tooling", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a8d5cf67c9fa01709bbfd7caa14007267bee5c7b"} +{"id": "train_44736", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3cd61dcde8ddac90c9012b716e630b6aef068dea"} +{"id": "train_44737", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fcafe368dcdaf134b995d8d3097f65b86e4d21fd"} +{"id": "train_44738", "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: 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", "reproducibility", "cost_latency_tradeoffs", "security_gates"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2cb0a5a05b4a3b3e90cd59b8cc898196beb13938"} +{"id": "train_44739", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "auditability", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "05f2e5111c20fdfc4d5b9adce9cafa6b1f8987ec"} +{"id": "train_44740", "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: 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", "auditability", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a76fdde5e91bae1bcaf4dd72adcc6cd849c5d0f5"} +{"id": "train_44741", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["reproducibility", "security_gates", "documentation", "evaluation_metrics"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a54f769a38a96e9716aa574e895f6fb28ee71d3"} +{"id": "train_44742", "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": ["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": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ad636e77bd282ccc8a233973407450ec46b2e406"} +{"id": "train_44743", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "documentation", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a6d35a01def86b063b511310cc4cd33f644a8b1d"} +{"id": "train_44744", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6ccf085b347f8773642c130e3f4ef80133d63f81"} +{"id": "train_44745", "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": ["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", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79723633f3177d9a6d70bcf9952a97a7ca0841f3"} +{"id": "train_44746", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "cost_latency_tradeoffs", "security_gates", "reproducibility"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "834763ac97c15740ca5111141a2a28fb20cf77fc"} +{"id": "train_44747", "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": ["documentation", "ci_integration", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69a4e68b1202f1950042bad8a596db7a8219896d"} +{"id": "train_44748", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "governance", "documentation", "tests_are_truth"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5dec9f1fa33d2ef11d0c26a173923c16a63bf3fd"} +{"id": "train_44749", "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: 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": ["tests_are_truth", "security_gates", "tooling", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "589fa069e9fc150bca03ff820d3f29a05972090a"} +{"id": "train_44750", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "66b7c32f77d8badc346e5a4847c182509d4f3dc3"} +{"id": "train_44751", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "tooling", "auditability", "documentation"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f50accf8b7edc0fd9e507b605ee048bcb3554750"} +{"id": "train_44752", "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: 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": ["tooling", "reproducibility", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2436f54fc0a76b17789f4fdc86030abfb689af12"} +{"id": "train_44753", "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: 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", "auditability", "security_gates", "ci_integration"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3dcd26293146240bfa733344753e1468d3da1585"} +{"id": "train_44754", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "reproducibility", "tests_are_truth"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cb0e31539393486e7e492b5eb774bd45974736ad"} +{"id": "train_44755", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "828c484128126477f799493949df963d0c4edbc7"} +{"id": "train_44756", "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", "cost_latency_tradeoffs", "repo_scale_reasoning", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb32dc2924139868ae3b62b26b5144b7a11c56f6"} +{"id": "train_44757", "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: 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": ["auditability", "tests_are_truth", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "39e262b8366b6097a489d83453352ae4403f3e15"} +{"id": "train_44758", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "tooling", "auditability"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6048d55dd2325401c3731cce61c3808e9d680bc6"} +{"id": "train_44759", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f8ac4447ff5f26177f3413935fc4962974bd6bb7"} +{"id": "train_44760", "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": ["reproducibility", "auditability", "tooling", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "299f02e5ec0fbbdc94e82f9ac010748f9433c0f7"} +{"id": "train_44761", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "ci_integration", "tests_are_truth"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7158c0bd4df402b8c8a1b99c5b215f6db20690ff"} +{"id": "train_44762", "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": ["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", "tooling", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b6cf117f3723a077de7855e5a35cb38985689f1"} +{"id": "train_44763", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "governance", "repo_scale_reasoning", "security_gates"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9fd0ab7592067a370e546a13c8e0da058a00c648"} +{"id": "train_44764", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "75abe0806a00ea6baddc25c3b7f167ae88e0e056"} +{"id": "train_44765", "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: 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", "ci_integration", "governance", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e6e5061201c8494720d142ce117da609e36bd196"} +{"id": "train_44766", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f9c74de10c606c16c5af90f45d889cc01b8e28a6"} +{"id": "train_44767", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "957048ba966c37be888d2251ad2e47c0bc804257"} +{"id": "train_44768", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4f5a7b4a4ad69f80b7e8b2e0c61d50dce2ca06fa"} +{"id": "train_44769", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling", "governance"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c9953914114815617660da3fba61f38abe397e6d"} +{"id": "train_44770", "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": ["tests_are_truth", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3316b30b10750cdff4d1412a061833baac29f78a"} +{"id": "train_44771", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "repo_scale_reasoning", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dd7d941bd9ee9011a3153f7917bdb3a1d6068194"} +{"id": "train_44772", "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: 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", "ci_integration", "tests_are_truth", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6916a45270bb4ffb50764397e60937e7e1a4097e"} +{"id": "train_44773", "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": ["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", "governance", "ci_integration"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a895e8916c1e99264903f5c72aafb94c2bb3aab2"} +{"id": "train_44774", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "governance"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "179bc2a19c75e8a36b293f43c0b40a4d4d7a6364"} +{"id": "train_44775", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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": ["evaluation_metrics", "repo_scale_reasoning", "auditability", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ceb36c8c7da5189252ca5247bd082a99b1e4b7aa"} +{"id": "train_44776", "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: 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", "security_gates", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "978e13b60d48e66063a7550384b75f361ee90ab4"} +{"id": "train_44777", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9af398ceb135ec07e106d90d4362a500f36c23a3"} +{"id": "train_44778", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "tooling", "tests_are_truth"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cbf5d3624f7df50b0dea73dd9d25d46d50b7320"} +{"id": "train_44779", "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: 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", "governance", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4200f0ffb83ee26e1aeb1711332a7c46f4b318b9"} +{"id": "train_44780", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "documentation", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b25f6a851ba7e950eb6d743baf99a4543a2813e7"} +{"id": "train_44781", "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: 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", "reproducibility", "documentation", "governance"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cd8d3c195ee424e06fe2bf53a0b611d0a4ca6d9d"} +{"id": "train_44782", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "documentation", "security_gates"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6cf2e1c12578e20d306aba5139d63cdd7119de36"} +{"id": "train_44783", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "auditability", "repo_scale_reasoning"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f58fe1b6ce0653e4dcf81c444e15d46a92f97960"} +{"id": "train_44784", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "security_gates", "governance", "documentation"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "16791bbb3017908aa7221f9d97a896d3e425ef46"} +{"id": "train_44785", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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": ["repo_scale_reasoning", "security_gates", "ci_integration", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67ab2bddf4cc7f0103de16c5242c089505113c5e"} +{"id": "train_44786", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "52f222d6d3083c47a77c64e704c0b8fb2f6bf049"} +{"id": "train_44787", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b49eadb2b2c4f8b7a458d12e7f7f214683fef6b0"} +{"id": "train_44788", "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: 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": ["documentation", "reproducibility", "ci_integration", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a56ecc82fc3ea60da25d5826db0a52418e006984"} +{"id": "train_44789", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "auditability", "tooling"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "287cc57a7c507e0fb2d0638a588318ab0343dadb"} +{"id": "train_44790", "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", "tests_are_truth", "evaluation_metrics", "documentation"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "618f97f315e3fc62601c574a44e3111012f37b3a"} +{"id": "train_44791", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "tooling", "tests_are_truth"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53fbde23464e3fba37b3abb8041122df123c03d7"} +{"id": "train_44792", "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": ["tooling", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2750524268a79692ffe21dd8880276942ac8d15f"} +{"id": "train_44793", "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: 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", "auditability", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dff4180bbf372f40ca4376fff7143fd9e585d38f"} +{"id": "train_44794", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "documentation", "security_gates"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d3a3f3b2ff97fa6f8d590d6bf8eb8ee0a6e17841"} +{"id": "train_44795", "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: 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", "tests_are_truth", "governance"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6557acfb47645b40ef455135f05f1df66ba914cc"} +{"id": "train_44796", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "repo_scale_reasoning", "reproducibility", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2fc3a987c135b8a821435dc461a53faa75fc0888"} +{"id": "train_44797", "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: 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", "security_gates", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2436f54fc0a76b17789f4fdc86030abfb689af12"} +{"id": "train_44798", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "tooling", "security_gates"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23f12dd4dfdccfdf5a3e48edccb2ea91de88c391"} +{"id": "train_44799", "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": ["security_gates", "cost_latency_tradeoffs", "tooling", "tests_are_truth"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe3f6ed84e5dd71ca863a3dcf06155ce12faf239"} +{"id": "train_44800", "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": ["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", "tooling", "documentation"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dddf6f8cb68e0bb837a910298007933221d9fce0"} +{"id": "train_44801", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["ci_integration", "repo_scale_reasoning", "governance", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23502e1f6203d8167ded3e2d25b3a73a0c4b7277"} +{"id": "train_44802", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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": ["evaluation_metrics", "tests_are_truth", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe4641cf61da7a68ebe948213223163d32c6765c"} +{"id": "train_44803", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "repo_scale_reasoning", "reproducibility"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8982cb0f7eede2dece86533ab4b0848c1edcbbaf"} +{"id": "train_44804", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "reproducibility", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bcfa2d04a4f9828812e448650768c242bac5949"} +{"id": "train_44805", "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: 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", "ci_integration", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b91a1f501189619c384e03f6569df72277967225"} +{"id": "train_44806", "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": [], "metadata": {"target_language": "Python", "developer_needs": ["cost_latency_tradeoffs", "reproducibility", "security_gates", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d4ac61ddde2e1488d65decc675ff5505a2b05dbb"} +{"id": "train_44807", "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": ["cost_latency_tradeoffs", "repo_scale_reasoning", "governance", "reproducibility"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7e0f81b4f575c16e1fa34e80afe84021a3718c03"} +{"id": "train_44808", "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": ["reproducibility", "documentation", "ci_integration", "evaluation_metrics"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b4fd83a756d23be0ea909dd7efe40f3af4abaa8f"} +{"id": "train_44809", "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: 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": ["reproducibility", "security_gates", "governance", "ci_integration"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bfd923aa1eec938fe42a9022e5975c592ba1cda"} +{"id": "train_44810", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "reproducibility", "documentation", "tooling"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a5dbda55964bf749a21fb040a8ffef5d5310b57c"} +{"id": "train_44811", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8a41cde524e82ec46964b6134592daa045976084"} +{"id": "train_44812", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "ci_integration"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ab67c984015f632d66efb5cf8458147bca5a91e6"} +{"id": "train_44813", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "evaluation_metrics", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "25147ed019fb8d11477f367bfcc69dec24fc0418"} +{"id": "train_44814", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "93968e44e5c5bb64444ab797f4732c085d5f95cb"} +{"id": "train_44815", "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: 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": ["governance", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bfd2ecbbbca042eea311e87b4cae207e60ebfe80"} +{"id": "train_44816", "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: 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": ["security_gates", "cost_latency_tradeoffs", "auditability", "ci_integration"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ebb66e87afbb0500a942e54f05daca780f05dc83"} +{"id": "train_44817", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "evaluation_metrics", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "999c02f738252a99194bdb1bd003fe155cc2041a"} +{"id": "train_44818", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "evaluation_metrics", "tests_are_truth", "governance"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "40776be96a5d6ca2c27d4d704e736ffe97aa724a"} +{"id": "train_44819", "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": ["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", "security_gates", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cbe4d1d275039daa5f56fa043ca359b297627026"} +{"id": "train_44820", "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: 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": ["documentation", "repo_scale_reasoning", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5262173011e84c5cf85e41d72d559b55d727a9e4"} +{"id": "train_44821", "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": ["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", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a9ba0aa6425babcaefcc0ba030f0b5901b492653"} +{"id": "train_44822", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71b8d662b76c7393b6993e088a8d2e2a88ecf0c4"} +{"id": "train_44823", "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: 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", "auditability", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a24f6533de505f92178b7a36f2d1f985e3a4e1a"} +{"id": "train_44824", "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": ["tooling", "documentation", "governance", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "969b47d187e87cf0e2e1786d158ec4034523d8b2"} +{"id": "train_44825", "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: 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", "tooling", "reproducibility", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "57cfeb3f8e1c37e6af897d62a954b1c711c5041b"} +{"id": "train_44826", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- 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", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7759306d913f6b6779080ce60664ac132d359941"} +{"id": "train_44827", "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: 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": ["repo_scale_reasoning", "documentation", "ci_integration", "tooling"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a8f164d507b57deef34add8962b289ee8f41ea2"} +{"id": "train_44828", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "cost_latency_tradeoffs", "repo_scale_reasoning", "documentation"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1b595ee5d99478d3bda98c8a15f4b5f3772c16a2"} +{"id": "train_44829", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "evaluation_metrics", "documentation", "tooling"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3bd8e610cc98a9d7b85a4d01346af84afa3c483"} +{"id": "train_44830", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "472fe1f79c19e2695c0741619313d59f5ddc013e"} +{"id": "train_44831", "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": ["tests_are_truth", "ci_integration", "evaluation_metrics", "documentation"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ddf4968006b4f99972691440f615f3333d9c723"} +{"id": "train_44832", "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": ["documentation", "repo_scale_reasoning", "ci_integration", "governance"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23422ac6a94a4d86f98b54934196ba215ee346d3"} +{"id": "train_44833", "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": ["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", "auditability", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c2017d1f1ab3104f75c7c6f1b4a2d65dd8a11a0c"} +{"id": "train_44834", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["repo_scale_reasoning", "reproducibility", "tooling", "governance"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "13512fa131180ac60930360cdb4a1e7cf4483b42"} +{"id": "train_44835", "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": ["cost_latency_tradeoffs", "tests_are_truth", "security_gates", "evaluation_metrics"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3df60378e76127a9ecd773eecc575d50bfaea328"} +{"id": "train_44836", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["tooling", "cost_latency_tradeoffs", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2b54ba475728a6fcbc158dfd726bde9b1774e347"} +{"id": "train_44837", "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", "reproducibility", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "65c9aaadb6c1e73cf4dee1cd486de6122f0a4457"} +{"id": "train_44838", "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": ["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", "documentation"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1e8b8166b9e7a03fc87bcc24c99af3c414038fa"} +{"id": "train_44839", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["governance", "security_gates", "tests_are_truth", "tooling"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6fd705d82a35bf58f21c591b03c88a0734c13576"} +{"id": "train_44840", "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": ["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", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f1e8b8166b9e7a03fc87bcc24c99af3c414038fa"} +{"id": "train_44841", "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": ["repo_scale_reasoning", "ci_integration", "cost_latency_tradeoffs", "tests_are_truth"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ffb5af248f6d71fdc182b1995cd227b257e4599b"} +{"id": "train_44842", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "tooling", "governance", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4dfa6f26921698a392efd255f418b3146d1d2646"} +{"id": "train_44843", "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": ["reproducibility", "ci_integration", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c911c35f6adb480509adab50926a6f35964e35e0"} +{"id": "train_44844", "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: 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", "tooling", "security_gates", "tests_are_truth"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e98aac2429ef79086e2089089d2be044a00af625"} +{"id": "train_44845", "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: 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": ["evaluation_metrics", "documentation", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1e464a99b118e51154da8d6d7e728cefd6e66984"} +{"id": "train_44846", "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: 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": ["documentation", "cost_latency_tradeoffs", "tooling", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4967f9c5c562d570a2d6362ee4ebeb10326d8851"} +{"id": "train_44847", "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: 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": ["evaluation_metrics", "tooling", "governance", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61e916c6b463deaa45e535c50f67dad6c16a5479"} +{"id": "train_44848", "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": ["tooling", "cost_latency_tradeoffs", "ci_integration", "reproducibility"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eda12ac838e7a900555c369075f596f4d2bef784"} +{"id": "train_44849", "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: 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", "evaluation_metrics", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79b52a69073c8335a9a59f36375a773fdf790221"} +{"id": "train_44850", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "evaluation_metrics"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8c602db113e5ff6725ad62e3279150edc6615f99"} +{"id": "train_44851", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "security_gates", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "86ba050a3d434a28d7913126329a18ce603762bb"} +{"id": "train_44852", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "security_gates", "tooling", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "83cdd776448499657509bc3c9fe6f560c62bd2be"} +{"id": "train_44853", "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: 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": ["cost_latency_tradeoffs", "evaluation_metrics", "ci_integration", "auditability"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18d149e7d5119be4fdf9a2dc6764c9daed5cb0c7"} +{"id": "train_44854", "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: 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", "tooling", "governance", "ci_integration"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ddc98ceb9cf4cd76dc92b576584614b38a7864c"} +{"id": "train_44855", "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: 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", "reproducibility", "governance", "ci_integration"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67e6b3e447fa97bd00d8be81097f805a421f474b"} +{"id": "train_44856", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "reproducibility", "repo_scale_reasoning", "governance"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "981cf04e5e76c538f039a3a2ced8503d3db90d09"} +{"id": "train_44857", "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: 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": ["evaluation_metrics", "repo_scale_reasoning", "security_gates", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "18a55717117dcc7639d7c0fbe78ac4bb8ea313a5"} +{"id": "train_44858", "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: 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": ["security_gates", "repo_scale_reasoning", "tooling", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2706d69657bc71093ee48051c507349aaf2a40c8"} +{"id": "train_44859", "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": ["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", "ci_integration", "governance"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b360361ef83ab5ddefdbefa10b8e83892d0f42a5"} +{"id": "train_44860", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "governance", "tooling", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fdc06aa1eaf1bec5750c2501758994e86aea3504"} +{"id": "train_44861", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "governance", "evaluation_metrics"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c64bb68a97dc5c53d9f44faf55ee1773a9099c3e"} +{"id": "train_44862", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "evaluation_metrics", "tooling"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9807c9d7a5539ac973a0c866748f1649d911c86e"} +{"id": "train_44863", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "governance", "security_gates", "reproducibility"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b7f930fdee283fd57890f75bd13384438ae81f2c"} +{"id": "train_44864", "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: 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", "governance", "repo_scale_reasoning", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "44b84ac93e0855fe899ba18f7e7cd3a0dc3813fe"} +{"id": "train_44865", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "ci_integration", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["governance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7be1ac5ce4abe9971c77becab0111cdbc5025613"} +{"id": "train_44866", "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: 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": ["tests_are_truth", "auditability", "governance", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "29127025b92ac490f9652f78d42a8acc190f7b5a"} +{"id": "train_44867", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a12d65180fdafe8044e134d1526e07fa29ced454"} +{"id": "train_44868", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "evaluation_metrics", "repo_scale_reasoning"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7641687be789e7ff0cd88f8fcf5c6af2b87665f8"} +{"id": "train_44869", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d9e355a5a9600c4b053f02fdb0070426237b5652"} +{"id": "train_44870", "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": ["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", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb104cd57514bcdc485c743d20dd5d706b6fae5f"} +{"id": "train_44871", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "documentation"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4958f5c74a9b17b66ef99b79c8c1467eba9e758e"} +{"id": "train_44872", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "auditability", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4a715f6993037b22367662fd542125b0a960fce2"} +{"id": "train_44873", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "documentation", "governance", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1ae78269b8c7551fb70cac8c547462cdd6d9efca"} +{"id": "train_44874", "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: 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": ["documentation", "tests_are_truth", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b1ac6c0b89c36400037259c7168245f7358c4a79"} +{"id": "train_44875", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51a11675b4682715d495495f667c06d81c945e1c"} +{"id": "train_44876", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "security_gates", "repo_scale_reasoning", "governance"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "84603585ce645f45cafc952f9bacaaa31c2e7d4e"} +{"id": "train_44877", "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: 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": ["repo_scale_reasoning", "tooling", "evaluation_metrics", "security_gates"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7dcdddc36fba5e06e044e6f986d73feadb4e2939"} +{"id": "train_44878", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "ci_integration", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ecaac099507841a0eda3fea80ac3ccf9066ccbcf"} +{"id": "train_44879", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "reproducibility"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2765f7c863be87c10aebb5aabcf65f245720c341"} +{"id": "train_44880", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71c6d7b174ca3b4cca464e51c0171ccef860ae5b"} +{"id": "train_44881", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "evaluation_metrics"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4ca3bb264c62bfb69b28a29debdace1a9603849f"} +{"id": "train_44882", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["governance", "tests_are_truth", "security_gates", "documentation"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f72bffdf6d43089278ebd63d7e6f8088002953ae"} +{"id": "train_44883", "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": ["tests_are_truth", "tooling", "security_gates", "governance"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0eca96a5362f57683f2078090c75292e7eee95f"} +{"id": "train_44884", "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": ["security_gates", "evaluation_metrics", "documentation", "tests_are_truth"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70fffef3fe1e848984682ee96998bbfcc46e4539"} +{"id": "train_44885", "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": ["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", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "70826dd52247054967a20032a1169505a6019b19"} +{"id": "train_44886", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "reproducibility"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "95de2a4fec0d93290a2889cf575a98820e432e5b"} +{"id": "train_44887", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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": ["documentation", "repo_scale_reasoning", "auditability", "cost_latency_tradeoffs"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2546857f7183fdb516c085231eae90e60d5c66d7"} +{"id": "train_44888", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "reproducibility", "auditability", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b2a9fb21cfcfdf13e5338c147541f74e5cb09946"} +{"id": "train_44889", "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: 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": ["evaluation_metrics", "documentation", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7310ef5138590bf7838b262b541610cb4fc1b85a"} +{"id": "train_44890", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "governance", "tooling"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3d9906b4a68c915eaad14648fe14c3b8ed01f25e"} +{"id": "train_44891", "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": ["repo_scale_reasoning", "security_gates", "auditability", "governance"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7a203f7ff8176ffd4bfbaa3a0a1dc19cbc019b92"} +{"id": "train_44892", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tooling", "auditability", "ci_integration"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2464f7265b27cd7d9174575eb44d07525ebd5e2f"} +{"id": "train_44893", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tooling", "governance", "evaluation_metrics", "documentation"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4cbfe77fd04e38108d7d445c4a53d0554f6d8182"} +{"id": "train_44894", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "auditability", "tests_are_truth", "tooling"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff3ce0b291ff3c20f4cd117349aa22a45ce14a8b"} +{"id": "train_44895", "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: 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", "tooling", "cost_latency_tradeoffs", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "71417bf1d95ba28ce3bb3106d3fbd119b7aa45c9"} +{"id": "train_44896", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "auditability", "documentation", "tooling"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "89a18bd6c93ee44490204485f3018d1046f3376c"} +{"id": "train_44897", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "reproducibility", "evaluation_metrics"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eadeb0945349df0f6d57f71125011f339ba95209"} +{"id": "train_44898", "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: 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": ["reproducibility", "cost_latency_tradeoffs", "documentation", "security_gates"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "268e41bc8fbb5397824dc23c175066f15cc2b98c"} +{"id": "train_44899", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "cost_latency_tradeoffs", "reproducibility", "documentation"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "98dfce3a6aec6a60fa9f54d8e900c1f60f40b618"} +{"id": "train_44900", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "reproducibility", "evaluation_metrics", "governance"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "01144fd0b5aba05b7af9d58fa9cb55394493115c"} +{"id": "train_44901", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tooling", "cost_latency_tradeoffs", "governance"], "moe_experts": ["agentic_systems_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b720c012383d00a7d0fa37176a85144d2c441f6a"} +{"id": "train_44902", "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: 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": ["auditability", "reproducibility", "documentation", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "55c8a16f109c277b4cdc19ed2571a079e1a7dfa9"} +{"id": "train_44903", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61a736c871929fa2644bd9bb79f40fb726aa8347"} +{"id": "train_44904", "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: 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", "security_gates", "tests_are_truth", "documentation"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "64001d8adeae529bed5b33e0103cd0ca8d6fcaf8"} +{"id": "train_44905", "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: 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", "repo_scale_reasoning", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "450ff3e9c51be3e91020ade42baa233e78c84641"} +{"id": "train_44906", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration"], "moe_experts": ["security_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3bfdd87d34562e801cc3c4713b8625d80ab073d9"} +{"id": "train_44907", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["security_gates", "reproducibility", "evaluation_metrics", "tests_are_truth"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daa46feb0024e4e8e04c20a5b5ff2d7727c3212e"} +{"id": "train_44908", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "69ec6251d327eac16e427befb43542a50d30ef4e"} +{"id": "train_44909", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2dea667d415a196663816053d2308584beb1f7b2"} +{"id": "train_44910", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "79a4eedc8990ced7024192c29e21330fe3a14b3b"} +{"id": "train_44911", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "evaluation_metrics", "governance", "documentation"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "06dd373dca09bc6937980c68d7e6ff0d547dcfe2"} +{"id": "train_44912", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "documentation", "evaluation_metrics", "security_gates"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "defff092d880df35613fa964cb45994a7cf43dc5"} +{"id": "train_44913", "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: 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": ["tests_are_truth", "repo_scale_reasoning", "reproducibility", "auditability"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "effe54370f24e8a826d6834ecae516ff9488ca80"} +{"id": "train_44914", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "evaluation_metrics", "auditability", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "ff45fae1a2252d8fc488ded5e6c57e7fd835bd1f"} +{"id": "train_44915", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51599702f7b85ded1c2dae6f4d04626928ce96de"} +{"id": "train_44916", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates"], "moe_experts": ["security_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b9ed9ca7f60e906f0c1845274f3654ae9a389a8"} +{"id": "train_44917", "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: 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", "governance", "tooling", "ci_integration"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8aea59f6a0e7699dafe1dd8ce240f07cc89bf971"} +{"id": "train_44918", "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": ["auditability", "repo_scale_reasoning", "ci_integration", "security_gates"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4bf30987a0da1d737b1a306ae57a510e3303c3d6"} +{"id": "train_44919", "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: 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", "ci_integration", "documentation", "reproducibility"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9092493249bf5ad15d4efbc2bba2982a55c54031"} +{"id": "train_44920", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "auditability", "tooling"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "daf747af4aa36c2df561501751c0f15af8499fac"} +{"id": "train_44921", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "cost_latency_tradeoffs", "tests_are_truth", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d06b78a076fb158cd966d868aff7f0ee3a677508"} +{"id": "train_44922", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "ci_integration", "tooling"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0c3c296c0e568359aa3062ee46bf327e0fa1b759"} +{"id": "train_44923", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "governance"], "moe_experts": ["governance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "fe5375551eb42c69523d48b79406df3adef15e85"} +{"id": "train_44924", "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: 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": ["tooling", "repo_scale_reasoning", "cost_latency_tradeoffs", "evaluation_metrics"], "moe_experts": ["performance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "08113a9c9317c8a730e1e9eacff4322450d09419"} +{"id": "train_44925", "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: 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": ["tooling", "tests_are_truth", "repo_scale_reasoning", "auditability"], "moe_experts": ["agentic_systems_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "8298127615672d21875221d832a2bc08d2b51753"} +{"id": "train_44926", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "reproducibility", "repo_scale_reasoning", "security_gates"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "018f1fa145a75af5831f7622c6146a06f7fd8575"} +{"id": "train_44927", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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": ["tooling", "tests_are_truth", "governance", "repo_scale_reasoning"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7581fb642a66aea84b3e47ff12f065277a1fbf29"} +{"id": "train_44928", "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: 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": ["tests_are_truth", "cost_latency_tradeoffs", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "5572941684c3d53d9832f273f14e63b241777845"} +{"id": "train_44929", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "evaluation_metrics", "tooling", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0e001aed2dc97cf50ac4d102f9ce5015b14e835c"} +{"id": "train_44930", "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": [], "metadata": {"target_language": "C#", "developer_needs": ["tests_are_truth", "repo_scale_reasoning", "documentation", "reproducibility"], "moe_experts": ["performance_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9d5b19930f3c6fd9d6a6b0f283e8e96567d081a9"} +{"id": "train_44931", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "governance", "auditability", "ci_integration"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "51a5c1d02226c4af77402f9018f5d1f23034ff1f"} +{"id": "train_44932", "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: 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", "documentation", "reproducibility", "tests_are_truth"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "77d686e2b208c7f3289d2b734fc5bd81f1fe0d44"} +{"id": "train_44933", "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": ["evaluation_metrics", "reproducibility", "auditability", "ci_integration"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "53f12c8e561269b168b877e735a05266430aafb1"} +{"id": "train_44934", "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: 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", "documentation", "cost_latency_tradeoffs", "governance"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "407b9fb72debba701439306af977e0718726e2b2"} +{"id": "train_44935", "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": [], "metadata": {"target_language": "Bash", "developer_needs": ["security_gates", "auditability", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c3408143855076a1a9949e121b5a874c415e4d36"} +{"id": "train_44936", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "tooling", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "c834ee9cb17ae5f13435905c143471e512145b13"} +{"id": "train_44937", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "330abc29e723d1502b5ddbc9f1155f07a42e600e"} +{"id": "train_44938", "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: 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": ["documentation", "tests_are_truth", "ci_integration", "reproducibility"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "23f388046b69f6d5aa1b418f7c255c0699485f51"} +{"id": "train_44939", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "documentation", "tests_are_truth", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6b30728a03c77a55ee2c129d143beac08f9edcde"} +{"id": "train_44940", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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", "auditability", "tooling", "ci_integration"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d84fbf494bab2bc739553328da3b22698ad13448"} +{"id": "train_44941", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "documentation", "tests_are_truth", "tooling"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "088d3d8bca008979df8a55d8a56c3006594882b8"} +{"id": "train_44942", "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": ["security_gates", "governance", "ci_integration", "repo_scale_reasoning"], "moe_experts": ["performance_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2555610479d82f75196908f0d1eab2df0d4ed0f7"} +{"id": "train_44943", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "documentation", "tests_are_truth", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "179bc2a19c75e8a36b293f43c0b40a4d4d7a6364"} +{"id": "train_44944", "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", "tooling", "governance", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "72d128472188ff21e6f286c6dce5817357b3637e"} +{"id": "train_44945", "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: 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", "documentation", "ci_integration", "auditability"], "moe_experts": ["agentic_systems_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d0ed46d40b1e34730fbee048f0138089e438d93b"} +{"id": "train_44946", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "auditability", "security_gates"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "1399b2101ee1ebada92e5a99a23acd83b4c746ea"} +{"id": "train_44947", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "tooling"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "da65526316801e470bba5adff4bc6767a04dff90"} +{"id": "train_44948", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "security_gates", "repo_scale_reasoning"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7b25224a2827f151d2e63b00f4a4bef618be25e7"} +{"id": "train_44949", "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: 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": ["repo_scale_reasoning", "auditability", "reproducibility", "tooling"], "moe_experts": ["security_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e61fd056df6ba461d90f3627ae57eac822d9eedb"} +{"id": "train_44950", "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: 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", "governance", "security_gates", "documentation"], "moe_experts": ["data_curation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "67e2c1b9d9b9874db44afb942531e91f7d7b0f52"} +{"id": "train_44951", "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": ["ci_integration", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["governance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9ad836ed42b6dc637d8211ec3a198a01ff54ce0e"} +{"id": "train_44952", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform 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", "tooling", "documentation"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b5682accf5a1ebf2b7ad011132f3c90db65eb52e"} +{"id": "train_44953", "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": ["security_gates", "evaluation_metrics", "ci_integration", "documentation"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b8409caf286bd3717b5b597b86949272225a92fa"} +{"id": "train_44954", "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: 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": ["repo_scale_reasoning", "ci_integration", "governance", "tooling"], "moe_experts": ["agentic_systems_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e1be7608221ac0dfe29b74578c995f3c651384e4"} +{"id": "train_44955", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "repo_scale_reasoning", "evaluation_metrics"], "moe_experts": ["coding_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0a925bfbe74ea9d6749381357591f4982201a6c6"} +{"id": "train_44956", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding 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", "ci_integration", "governance", "security_gates"], "moe_experts": ["data_curation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "310a6f83542ec275108bf8e6d490e2343efb3a87"} +{"id": "train_44957", "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: Large monorepo with flaky tests and strict CI.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "tests_are_truth", "security_gates", "ci_integration"], "moe_experts": ["coding_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "68af15e343752f5b5409014e911cc0fd765a3ae8"} +{"id": "train_44958", "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", "ci_integration", "tests_are_truth", "evaluation_metrics"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2ee1e0b444da60df9208d10b361605b76cc573b9"} +{"id": "train_44959", "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": ["reproducibility", "governance", "documentation", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "881709ce6edf1ec4bc57821e77801654b0e44853"} +{"id": "train_44960", "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": ["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", "ci_integration"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a186e87ba5b6351d53019a599bc1d7626f1b6f93"} +{"id": "train_44961", "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": ["documentation", "cost_latency_tradeoffs", "security_gates", "governance"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bb2ff79bf0d5c2b1a0d2db702d8fde183fdae6ed"} +{"id": "train_44962", "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: 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": ["evaluation_metrics", "auditability", "tests_are_truth", "governance"], "moe_experts": ["performance_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "4e7ebda4d08bce3a035a13958c296eb13405ab67"} +{"id": "train_44963", "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", "cost_latency_tradeoffs", "repo_scale_reasoning"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "3032205c6def0d1b97243fda49547ac462997167"} +{"id": "train_44964", "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": ["governance", "repo_scale_reasoning", "auditability", "security_gates"], "moe_experts": ["agentic_systems_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9511e8cd50a947ac8c2a2774e748c913e4e908ad"} +{"id": "train_44965", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "tests_are_truth", "governance"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "bf8e3d344cca7cdbebdb35ccc218f5a068de4e32"} +{"id": "train_44966", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["auditability", "tooling", "tests_are_truth", "ci_integration"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9762443951d03bb05c2ce971e259b7d96bf59d05"} +{"id": "train_44967", "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: 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", "reproducibility", "repo_scale_reasoning", "auditability"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "9cd1e5d14987df743152de1f33787452a481d826"} +{"id": "train_44968", "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: 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", "documentation", "tests_are_truth", "ci_integration"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "f6c5d00ef1aee2a9f395458cbaa522af2a832de6"} +{"id": "train_44969", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "ci_integration", "repo_scale_reasoning", "auditability"], "moe_experts": ["security_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "10b13d3982bd49841e2d74bcd935cc78074dda05"} +{"id": "train_44970", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "governance", "tests_are_truth"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "21df9a7600330da9ef888552dbc0e59e0d38acca"} +{"id": "train_44971", "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": ["repo_scale_reasoning", "reproducibility", "governance", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "628156ae0be9347d2e8d664e126abd3870767a7f"} +{"id": "train_44972", "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: 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": ["cost_latency_tradeoffs", "security_gates", "reproducibility", "ci_integration"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "082296cc5bd92efebd189dcce3be5ea9e0e52335"} +{"id": "train_44973", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 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", "security_gates", "tooling", "auditability"], "moe_experts": ["agentic_systems_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "261c35d60100851c762f57b8cee8b79ca12dc61d"} +{"id": "train_44974", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "documentation", "governance", "tests_are_truth"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d7e2a7cce9b6b3920bc67792c57751aad8113600"} +{"id": "train_44975", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "auditability", "tests_are_truth"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "6a2329a8b2a8cad0197adb7b12bc097ea75311f6"} +{"id": "train_44976", "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: 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", "ci_integration", "tooling", "evaluation_metrics"], "moe_experts": ["data_curation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d1f00e37780ddaccaafd78bcc3ced7f46d1bf072"} +{"id": "train_44977", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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", "governance", "ci_integration", "evaluation_metrics"], "moe_experts": ["coding_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dbd4050fa7027b516d6ab94fcaaf383491c00a3d"} +{"id": "train_44978", "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: Research team validating claims against real repos.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and 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", "security_gates", "ci_integration", "governance"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "cac6686406ca55ae8225d3b438fe53719568998c"} +{"id": "train_44979", "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": ["evaluation_metrics", "cost_latency_tradeoffs", "repo_scale_reasoning", "auditability"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e2003591781cf3aa13cc8b7958ea7292b8c65ced"} +{"id": "train_44980", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tests_are_truth", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["data_curation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "dc69c5fd5cfe84a83c134ee18fea0975846f0b07"} +{"id": "train_44981", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, 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", "tests_are_truth", "reproducibility", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "20be21bdde3dcd2e712879ac36ab4f4c8938db84"} +{"id": "train_44982", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nFailure:\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", "reproducibility", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["coding_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7f09dbe123b5eacffe5a959dea106611b9b8b4c6"} +{"id": "train_44983", "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: 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", "reproducibility", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "e349b3de38c77b62dc52c3890cb5a2e55a3b4232"} +{"id": "train_44984", "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: Regulated environment requiring audit trails.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nPatch (diff-style):\n```diff\n- if x == 0:\n- return 1/x\n+ 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": ["tooling", "governance", "security_gates", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "7587c5b0ec9bb3ccb00b1b844292aa6dd238e4e0"} +{"id": "train_44985", "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": [], "metadata": {"target_language": "TypeScript", "developer_needs": ["auditability", "tests_are_truth", "repo_scale_reasoning", "cost_latency_tradeoffs"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b79531042cf95a55db5ad90bd49d8a642aed2686"} +{"id": "train_44986", "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: Offline/local deployment with limited compute.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass generation.\n\nDesign with risks, metrics, acceptance 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", "documentation", "evaluation_metrics", "reproducibility"], "moe_experts": ["governance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "d674d3723c9cec1cc800efdef8cded169ed3954e"} +{"id": "train_44987", "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": [], "metadata": {"target_language": "Java", "developer_needs": ["tooling", "security_gates", "tests_are_truth", "repo_scale_reasoning"], "moe_experts": ["security_expert", "governance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "27beaaeeab855d2314c36ff885a62c77f792b645"} +{"id": "train_44988", "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": [], "metadata": {"target_language": "JavaScript", "developer_needs": ["ci_integration", "evaluation_metrics", "documentation", "reproducibility"], "moe_experts": ["coding_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2f30244dd7159d32d78242629de68a434cd4aad0"} +{"id": "train_44989", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with 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": ["reproducibility", "evaluation_metrics", "cost_latency_tradeoffs", "tooling"], "moe_experts": ["evaluation_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "b3708d9ea50ddc19ac9889af380b53ca52e07d7d"} +{"id": "train_44990", "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": [], "metadata": {"target_language": "Rust", "developer_needs": ["reproducibility", "security_gates", "evaluation_metrics", "governance"], "moe_experts": ["governance_expert", "agentic_systems_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0ddf2f41c5b1bef16a6a862b8f8452edd611f618"} +{"id": "train_44991", "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": ["tests_are_truth", "documentation", "auditability", "ci_integration"], "moe_experts": ["performance_expert", "security_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "61a7ead3504ffe27225c374f9246ec4050cb6826"} +{"id": "train_44992", "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: 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": ["reproducibility", "auditability", "evaluation_metrics", "documentation"], "moe_experts": ["coding_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "2e937da406824750b4370ddf44235d948f165175"} +{"id": "train_44993", "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": [], "metadata": {"target_language": "Go", "developer_needs": ["tooling", "repo_scale_reasoning", "reproducibility", "cost_latency_tradeoffs"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "94b13a16538b57baf55c0634176a9013dc5edbfe"} +{"id": "train_44994", "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: 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": ["reproducibility", "governance", "tests_are_truth", "tooling"], "moe_experts": ["evaluation_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "eb10dd3d892dcfae7c2505d6f8f2e7011cae3016"} +{"id": "train_44995", "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": ["reproducibility", "repo_scale_reasoning", "governance", "security_gates"], "moe_experts": ["governance_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "a291b431b91773f292c421f4c35cf99749b71b3f"} +{"id": "train_44996", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic 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", "cost_latency_tradeoffs", "auditability", "tooling"], "moe_experts": ["evaluation_expert", "data_curation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "0646cea336de9459a3a4ee16cec0a63fad745ff8"} +{"id": "train_44997", "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: 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", "repo_scale_reasoning", "cost_latency_tradeoffs", "auditability"], "moe_experts": ["data_curation_expert", "evaluation_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "447f6e644725c7495f1624061d0e708da88e51cf"} +{"id": "train_44998", "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: High-traffic service with latency SLOs.\nProduce expert-level, production-ready artifacts.", "input": "", "output": "Facts:\n- Modern AI coding prioritizes correctness, evaluation, and governance.\n- Agentic loops with test gates outperform single-pass 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", "tooling", "auditability", "tests_are_truth"], "moe_experts": ["evaluation_expert", "performance_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "671048bba726ec6cd33467cf760e51a5b0e34961"} +{"id": "train_44999", "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: 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", "security_gates", "evaluation_metrics", "cost_latency_tradeoffs"], "moe_experts": ["performance_expert", "coding_expert"], "governance": {"audit_required": true, "tests_required": true}}, "hash": "87e0363b144ae08a48476f3c549e0a84643d828d"}